A method for generating nursing plans for lung cancer patients at home based on adaptive neuro-fuzzy reasoning

Abstract

The present invention discloses a method for generating nursing plans for home-based lung cancer patients based on adaptive neural fuzzy reasoning, including S1, a comprehensive health data set; S2, in-depth analysis of the comprehensive health data set through an adaptive neural fuzzy reasoning model; S3, generating a nursing plan based on the in-depth analysis result of step S2; S4, collecting and analyzing changes in physiological parameters in the comprehensive health data set, self-feedback of patients, and changes in environmental factors; S5, adaptively adjusting the nursing plan based on the data collected in step S4, and adapting to changes in the patient's status by updating the nursing plan; S6, enabling medical professionals to remotely monitor the patient's status and provide feedback and suggestions to the patient through telemedicine technology. The present invention can analyze the comprehensive health data set of home-based lung cancer patients in real time by using an adaptive neural fuzzy reasoning system to generate a reasonable nursing plan.

Description

Technical Field

[0001] The present invention relates to the technical field of lung cancer nursing, and in particular to a method for generating a nursing plan for lung cancer patients at home based on adaptive neural fuzzy reasoning. Background Art

[0002] In the modern medical field, especially in the management of chronic diseases such as lung cancer, the application of home care and remote monitoring technologies is becoming more and more common. These technologies aim to provide patients with continuous health management and support through continuous monitoring and real-time data analysis, thereby optimizing treatment outcomes and improving quality of life. Existing technologies include the use of wearable devices to collect physiological data (such as heart rate, blood pressure, and blood oxygen saturation), as well as the collection of self-reported and lifestyle information from patients through mobile applications.

[0003] Nevertheless, existing technologies are clearly deficient in several key areas. First, existing home care systems are generally not flexible enough to adjust care plans in real time according to changes in the patient's condition. These systems often rely on fixed algorithms and preset parameters, and are difficult to adapt to the rapid changes in the health status of individual patients or their personalized needs.

[0004] Second, despite the collection of abundant health data, existing technologies still face challenges in comprehensively utilizing these data to generate or adjust care plans. The fusion processing, analysis and application of multi-source data in care decision-making are often insufficient, which limits the effectiveness and adaptability of care plans. In addition, the handling of uncertainty and ambiguity in the data processing process is not sophisticated enough, resulting in care recommendations that may not be precise or applicable.

[0005] Furthermore, although environmental factors have a significant impact on the health status of lung cancer patients, existing technologies often ignore the monitoring and analysis of environmental parameters, such as air quality, temperature and humidity. If changes in these environmental factors are not reflected in the adjustment of care plans in a timely manner, they may have an adverse impact on the health of patients.

[0006] Therefore, how to provide a method for generating nursing plans for lung cancer patients at home based on adaptive neural fuzzy reasoning is an urgent problem to be solved by those skilled in the art. Summary of the invention

[0007] One object of the present invention is to propose a method for generating nursing plans for home-based lung cancer patients based on adaptive neuro-fuzzy reasoning. The present invention can analyze the comprehensive health data set of home-based lung cancer patients in real time by using an adaptive neuro-fuzzy reasoning system. The analysis is not only based on static health conditions, but is continuously monitored and updated according to the dynamic changes of the patient's condition, so as to generate a reasonable nursing plan.

[0008] A method for generating a nursing plan for lung cancer patients at home based on adaptive neural fuzzy reasoning according to an embodiment of the present invention comprises the following steps:

[0009] S1. Collect the heart rate, blood pressure, blood oxygen saturation, sleep quality, activity level, eating habits and adverse reactions data of chemotherapy, immunotherapy and targeted therapy during treatment of lung cancer patients at home to build a comprehensive health data set;

[0010] S2. performing an in-depth analysis of the comprehensive health data set by using an adaptive neural-fuzzy inference model, wherein the adaptive neural-fuzzy inference model combines the learning ability of a neural network with the reasoning ability of fuzzy logic to process health data containing uncertainty and ambiguity;

[0011] S3. Generate a nursing plan based on the in-depth analysis results of step S2, including treatment suggestions, lifestyle adjustment guidance and health monitoring plan;

[0012] S4. Use a dynamic feedback mechanism to continuously monitor the patient's health status and the implementation of the nursing plan, collect and analyze changes in physiological parameters, patient self-feedback, and changes in environmental factors in the comprehensive health data set;

[0013] S5. Adaptively adjust the nursing plan based on the data collected in step S4, and adapt to changes in the patient's condition by updating the nursing plan;

[0014] S6. Through telemedicine technology, medical professionals can remotely monitor the patient's condition and provide feedback and advice to the patient.

[0015] Optionally, the S1 includes the following steps:

[0016] S11. Collect physiological parameters of lung cancer patients at home, continuously collect heart rate HR data through ECG monitoring equipment, regularly record blood pressure BP data using an automatic sphygmomanometer, and monitor blood oxygen saturation SpO2 in real time through a pulse oximeter. When collecting data, each parameter is recorded at a frequency of one minute;

[0017] S12. Automatically collect and analyze the patient's sleep quality data SQ and activity data AL, wherein the sleep quality data is monitored by a sleep monitoring device throughout the sleep cycle to evaluate the patient's sleep stage, duration, and interruptions:

[0018] SQ=α×REM+β×NREM+γ×AWAKE;

[0019] Among them, REM, NREM and AWAKE represent the duration ratio of rapid eye movement sleep, non-rapid eye movement sleep and wakefulness, respectively, and α, β and γ are weight coefficients, which are adjusted to optimize the sleep quality score;

[0020] Activity data AL records the patient's steps, exercise type, and duration by syncing with wearable health trackers;

[0021] S13. Record the patient's eating habits DH. Collect the type, amount and time of daily diet through the interaction between the patient and the nutrition management application. At the same time, analyze the diet photos uploaded by the patient through image recognition technology to automatically identify food categories and estimate intake:

[0022]

[0023] Among them, C j represents the intake of the jth food, ω j is the nutritional weight of the corresponding food, M is the total number of food categories;

[0024] S14. Systematically collect adverse reaction data TR of patients during treatment, including adverse reactions of chemotherapy, immunotherapy, and targeted therapy. Automatically update the data through the electronic adverse reaction record form filled out by the patient, and analyze it in combination with the changes in the patient's physiological parameters to identify the correlation between possible adverse reactions and specific treatments;

[0025] S15, comprehensively process and standardize the collected multi-source health data, including heart rate HR, blood pressure BP, blood oxygen saturation SpO2, sleep quality data SQ, activity data AL, eating habits DH and adverse reaction data TR, to construct a comprehensive health data set HDS:

[0026] HDS={HR, BP, SpO2, SQ, AL, DH, TR}.

[0027] Optionally, the S14 includes the following steps:

[0028] S141. Use the electronic health record system to automatically record the time series data of the heart rate HR, blood pressure BP, and blood oxygen saturation SpO2 of patients after chemotherapy, immunotherapy, and targeted therapy. According to the relationship between the changes in physiological parameters and the treatment time point, calculate the physiological response index R after each treatment. Use the time-weighted average method and add an exponential decay factor to make the physiological response index R sensitive to the recent physiological changes after treatment:

[0029]

[0030] Where T is the length of time from the start of monitoring after treatment, α, β and γ are the weights of different physiological parameters, reflecting their sensitivity in adverse reaction detection, and Z is the normalization constant to keep R within a reasonable range:

[0031]

[0032] S142. Automatically update data through the electronic adverse reaction record form filled out by the patient, recording the frequency and intensity of adverse reactions including nausea, vomiting, hair loss, and fatigue;

[0033] S143. Combined with the physiological parameter response index R and the patient's self-reported adverse reaction data, a multivariate logistic regression model was used for analysis. Interaction terms and nonlinear terms were added to identify and predict the correlation between specific treatments and specific adverse reactions, and the patient's adverse reaction risk was assessed. The formula is as follows:

[0034]

[0035] Where P(y=1|x) is the probability of a specific adverse reaction occurring under a given treatment and physiological response index, β0, β j , β jk , γ j are model parameters, x j are predictor variables including treatment type, physiological response index R, and other related health indicators, x j x k Represents variable x j and x k The interaction term is used to model the interaction between two treatments or physiological responses. is the variable x j The quadratic term of is used to capture its nonlinear effect on the probability.

[0036] Optionally, S2 includes the following steps:

[0037] S21. The adaptive neural fuzzy inference system is used to process the comprehensive health data set HDS. The adaptive neural fuzzy inference system combines the adaptive learning ability of the neural network and the reasoning ability of fuzzy logic to process the uncertainty and ambiguity in the data. The specific steps include:

[0038] Use an independent fuzzification function fuzzify for each health parameter to generate the corresponding fuzzy set;

[0039] The fuzzified health parameter set is used as input to construct the input layer of the fuzzy inference system;

[0040] S22. The neural network component in the adaptive neural fuzzy reasoning system learns and trains the input fuzzy set, automatically extracts fuzzy rules, and adjusts the fuzzy logic rules and neural network weights to optimize the generation of nursing plans:

[0041]

[0042] Among them, Δw x is the weight update amount, corresponding to the health parameter x, μx (x j ) is the fuzzy membership of the health parameter x in the jth sample, is the derivative of the activation function, λ is the regularization coefficient, used for L1 regularization of the weights, n is the number of samples, g(net j ) is the activation function of the neural network, net j is the net input to the neuron, calculated as:

[0043] net j =∑ x∈X w x ·μ x (x j );

[0044] S23. Use the back propagation algorithm to train the neural network, optimize the weights of the fuzzy logic combination, and apply the error gradient calculation formula for each output node error:

[0045]

[0046] Among them, output k is the network output of the kth training sample, target k is the target output of the kth training sample, sigmoid′(net k ) is the derivative of the activation function, which calculates the net input at the kth sample, input ij is the i-th input connected to the j-th neuron, and N is the total number of training samples;

[0047] S24. Through steps S21-S23, the adaptive neuro-fuzzy inference model is dynamically adjusted according to the actual patient data to provide a nursing plan to cope with changes in the patient's health status.

[0048] Optionally, the S21 includes the following steps:

[0049] S211. For each health parameter x, including heart rate, blood pressure, blood oxygen saturation, sleep quality, activity level, eating habits, and adverse reaction data on chemotherapy, immunotherapy, and targeted therapy during treatment, a fuzzy set is generated through a predefined fuzzy function fuzzify(x). Each fuzzy function maps a real number input to the membership of a fuzzy set and uses a Gaussian function for fuzzification:

[0050]

[0051] Among them, μ x is the element x i The degree of membership to the fuzzy set x, c x and σ xare the center and standard deviation of the Gaussian function, respectively, which are preset according to the specific distribution characteristics of the health parameters;

[0052] S212. Design a neural network layer for each fuzzy set, receive the fuzzified input, make preliminary weight adjustments, and update the weights using the gradient descent method:

[0053]

[0054] Where η is the learning rate, E is the prediction error, and the weight w corresponds to the contribution of each health parameter in the fuzzy set;

[0055] S213, integrate all individual neural network layer outputs, merge information through the fusion layer, and fuse the outputs using a weighted summation method to form a unified output:

[0056]

[0057] Among them, output x is the output from the neural network layer corresponding to the health parameter x, w x is the corresponding weight, reflecting the importance of each health parameter in the final decision.

[0058] Optionally, S3 includes the following steps:

[0059] S31, receiving the output result of step S23, reflecting the patient's current health status and potential care needs;

[0060] S32. According to the specific value of V, a preset logic determination rule set is applied to determine the patient's treatment recommendations, lifestyle adjustment guidance, and health monitoring plan. The logic determination rule set R is implemented in the form of a multi-condition decision table, as follows:

[0061]

[0062] Among them, v1 and v2 are preset thresholds used to distinguish different levels of care needs;

[0063] S33. Generate specific nursing plans for each level of nursing needs:

[0064] The light care program includes basic health monitoring and regular assessment of lifestyle habits;

[0065] The moderate care program adds customized nutrition advice and moderate physical activity guidance;

[0066] Intensive care options include intensive health monitoring, medication management, and rehabilitation programs;

[0067] S34. Set corresponding monitoring plan parameters for each level of care plan. Parameters P include monitoring frequency, key health indicators and feedback mechanism:

[0068] P 轻度 =(F 月 , {HR, BP}, automatic feedback system);

[0069] P 中度 =(F 周 ,{HR, BP, SpO2, AL},Telemedicine support;

[0070] P 重度 =(F 日 , {HR, BP, SpO2, SQ, TR}, expert team intervention);

[0071] Among them, F represents the monitoring frequency, which is F per month. 月 、Weekly F 周 or daily F 日 ;

[0072] S35. The generated care plan and monitoring plan are automatically uploaded to the patient's electronic health record, and updated and notified in real time through the patient's mobile health application.

[0073] Optionally, S4 includes the following steps:

[0074] S41. A real-time data monitoring system is set up to continuously collect the patient's physiological parameters, self-feedback and environmental factors. The real-time data monitoring system uploads the patient's heart rate HR, blood pressure BP, blood oxygen saturation SpO2, and other feedback and environmental data collected through the patient's mobile device to the cloud database in real time;

[0075] S42. Regularly evaluate the comprehensive health data set HDS and calculate the change Δ of each monitoring indicator:

[0076] ΔX=X current -X previous ;

[0077] Among them, X represents any monitoring indicator, X current is the current monitoring value, X previous is the value of the last record;

[0078] S43. Analyze the collected data using the data analysis module to identify any significant health trends or anomalies. The module uses preset health parameter thresholds for anomaly detection:

[0079] If ΔHR>HR threshold orΔBP>BP threshold then trigger alert;

[0080] Among them, HR threshold and BP threshold These are preset thresholds for heart rate and blood pressure that trigger health alerts.

[0081] Optionally, S5 includes the following steps:

[0082] S51. Systematic assessment of the patient's current health status and changes in environmental factors, including analysis of changes in monitored physiological parameters and environmental parameters such as air quality and temperature;

[0083] S52. Use the decision support system to adaptively adjust the nursing plan. The system uses predefined adjustment rules to update the nursing plan according to changes in health status. The rules determine whether adjustment is needed based on the following logical expression:

[0084] Adjustment Required=(|ΔX|>T X )for any X∈{HR, BP, SpO2, Env};

[0085] Where ΔX is the change in a given health or environmental parameter, T X is the threshold of the parameter, which is used to determine whether the change is significant;

[0086] S53. Automatically update the patient's care plan based on the output of the decision support system, including adjusting drug dosage, treatment frequency, lifestyle recommendations, and environmental control measures;

[0087] S54, realize environmental factor adjustment, automatically adjust the living habit guidance and health monitoring plan according to the changes of the patient's living environment:

[0088] When the environmental monitoring system detects a decrease in air quality, it automatically increases the use guidelines and frequency of indoor air purifiers;

[0089] Air quality has deteriorated, increasing the guidelines and frequency of use of indoor air purifiers;

[0090] Temperature fluctuations: When the temperature is too high or too low, the temperature control device settings at home will be automatically adjusted, and wearable suggestions will be provided;

[0091] Humidity changes: adjust the use of dehumidifiers or humidifiers based on high or low humidity, and remind patients about respiratory care in high or low humidity environments.

[0092] The beneficial effects of the present invention are:

[0093] (1) The present invention can analyze the comprehensive health data set of home-based lung cancer patients in real time by using an adaptive neuro-fuzzy inference system. The analysis is not only based on static health conditions, but is continuously monitored and updated according to the dynamic changes in the patient's condition, thereby generating a reasonable care plan.

[0094] (2) With the help of real-time data monitoring and adaptive adjustment mechanisms, the present invention can effectively respond to any changes in the patient's condition. When significant changes in key health indicators (such as heart rate, blood pressure) or environmental factors (such as air quality, temperature) are monitored, the care plan can be automatically updated to ensure that the treatment recommendations are always consistent with the patient's latest health status.

[0095] (3) The present invention focuses on the impact of environmental factors on the health of lung cancer patients and incorporates them into health data analysis and nursing plan adjustment, including monitoring air quality, temperature and humidity, etc., and automatically adjusting nursing recommendations based on these parameters, which helps to create a living environment that is more suitable for the health of patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0096] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0097] Figure 1 This is a flow chart of a method for generating a nursing plan for lung cancer patients at home based on adaptive neuro-fuzzy reasoning proposed by the present invention. DETAILED DESCRIPTION

[0098] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0099] refer to Figure 1 , a method for generating nursing plans for lung cancer patients at home based on adaptive neuro-fuzzy reasoning, comprising the following steps:

[0100] S1. Collect the heart rate, blood pressure, blood oxygen saturation, sleep quality, activity level, eating habits and adverse reactions data of chemotherapy, immunotherapy and targeted therapy during treatment of lung cancer patients at home to build a comprehensive health data set;

[0101] In this implementation, S1 includes the following steps:

[0102] S11. Collect physiological parameters of lung cancer patients at home, continuously collect heart rate HR data through ECG monitoring equipment, regularly record blood pressure BP data using an automatic sphygmomanometer, and monitor blood oxygen saturation SpO2 in real time through a pulse oximeter. When collecting data, each parameter is recorded at a frequency of one minute;

[0103] S12. Automatically collect and analyze the patient's sleep quality data SQ and activity data AL, wherein the sleep quality data is monitored by a sleep monitoring device throughout the sleep cycle to evaluate the patient's sleep stage, duration, and interruptions:

[0104] SQ=α×REM+β×NREM+γ×AWAKE;

[0105] Among them, REM, NREM and AWAKE represent the duration ratio of rapid eye movement sleep, non-rapid eye movement sleep and wakefulness, respectively, and α, β and γ are weight coefficients, which are adjusted to optimize the sleep quality score;

[0106] Activity data AL records the patient's steps, exercise type, and duration by syncing with wearable health trackers;

[0107] S13. Record the patient's eating habits DH. Collect the type, amount and time of daily diet through the interaction between the patient and the nutrition management application. At the same time, analyze the diet photos uploaded by the patient through image recognition technology to automatically identify food categories and estimate intake:

[0108]

[0109] Among them, C j represents the intake of the jth food, ω j is the nutritional weight of the corresponding food, M is the total number of food categories;

[0110] S14. Systematically collect adverse reaction data TR of patients during treatment, including adverse reactions of chemotherapy, immunotherapy, and targeted therapy. Automatically update the data through the electronic adverse reaction record form filled out by the patient, and analyze it in combination with the changes in the patient's physiological parameters to identify the correlation between possible adverse reactions and specific treatments;

[0111] S15, comprehensively process and standardize the collected multi-source health data, including heart rate HR, blood pressure BP, blood oxygen saturation SpO2, sleep quality data SQ, activity data AL, eating habits DH and adverse reaction data TR, to construct a comprehensive health data set HDS:

[0112] HDS={HR, BP, SpO2, SQ, AL, DH, TR}.

[0113] In this implementation, S14 includes the following steps:

[0114] S141. Use the electronic health record system to automatically record the time series data of the heart rate HR, blood pressure BP, and blood oxygen saturation SpO2 of patients after chemotherapy, immunotherapy, and targeted therapy. According to the relationship between the changes in physiological parameters and the treatment time point, calculate the physiological response index R after each treatment. Use the time-weighted average method and add an exponential decay factor to make the physiological response index R sensitive to the recent physiological changes after treatment:

[0115]

[0116] Where T is the length of time from the start of monitoring after treatment, α, β and γ are the weights of different physiological parameters, reflecting their sensitivity in adverse reaction detection, and Z is the normalization constant to keep R within a reasonable range:

[0117]

[0118] S142. Automatically update data through the electronic adverse reaction record form filled out by the patient, recording the frequency and intensity of adverse reactions including nausea, vomiting, hair loss, and fatigue;

[0119] S143. Combined with the physiological parameter response index R and the patient's self-reported adverse reaction data, a multivariate logistic regression model was used for analysis. Interaction terms and nonlinear terms were added to identify and predict the correlation between specific treatments and specific adverse reactions, and the patient's adverse reaction risk was assessed. The formula is as follows:

[0120]

[0121] Where P(y=1|x) is the probability of a specific adverse reaction occurring under a given treatment and physiological response index, β0, β j , β jk , γ j are model parameters, x j are predictor variables including treatment type, physiological response index R, and other related health indicators, x j x k Represents variable x j and x k The interaction term is used to model the interaction between two treatments or physiological responses. is the variable x j The quadratic term of is used to capture its nonlinear effect on the probability.

[0122] S2. In-depth analysis of comprehensive health data sets through adaptive neuro-fuzzy inference models. The adaptive neuro-fuzzy inference model combines the learning ability of neural networks and the reasoning ability of fuzzy logic to process health data containing uncertainty and ambiguity.

[0123] In this implementation, S2 includes the following steps:

[0124] S21. The adaptive neural fuzzy inference system is used to process the comprehensive health data set HDS. The adaptive neural fuzzy inference system combines the adaptive learning ability of the neural network and the reasoning ability of fuzzy logic to process the uncertainty and ambiguity in the data. The specific steps include:

[0125] Use an independent fuzzification function fuzzify for each health parameter to generate the corresponding fuzzy set;

[0126] The fuzzified health parameter set is used as input to construct the input layer of the fuzzy inference system;

[0127] S22. The neural network component in the adaptive neural fuzzy reasoning system learns and trains the input fuzzy set, automatically extracts fuzzy rules, and adjusts the fuzzy logic rules and neural network weights to optimize the generation of nursing plans:

[0128]

[0129] Among them, Δw x is the weight update amount, corresponding to the health parameter x, μ x (x j ) is the fuzzy membership of the health parameter x in the jth sample, is the derivative of the activation function, λ is the regularization coefficient, used for L1 regularization of the weights, n is the number of samples, g(net j ) is the activation function of the neural network, net j is the net input to the neuron, calculated as:

[0130] net j =∑ x∈X w x ·μ x (x j );

[0131] S23. Use the back propagation algorithm to train the neural network, optimize the weights of the fuzzy logic combination, and apply the error gradient calculation formula for each output node error:

[0132]

[0133] Among them, output k is the network output of the kth training sample, target k is the target output of the kth training sample, sigmoid′(net k ) is the derivative of the activation function, which calculates the net input at the kth sample, input ijis the i-th input connected to the j-th neuron, and N is the total number of training samples;

[0134] S24. Through steps S21-S23, the adaptive neuro-fuzzy inference model is dynamically adjusted according to the actual patient data to provide a nursing plan to cope with changes in the patient's health status.

[0135] In this implementation, S21 includes the following steps:

[0136] S211. For each health parameter x, including heart rate, blood pressure, blood oxygen saturation, sleep quality, activity level, eating habits, and adverse reaction data on chemotherapy, immunotherapy, and targeted therapy during treatment, a fuzzy set is generated through a predefined fuzzy function fuzzify(x). Each fuzzy function maps a real number input to the membership of a fuzzy set and uses a Gaussian function for fuzzification:

[0137]

[0138] Among them, μ x is the element x i The degree of membership to the fuzzy set x, c x and σ x are the center and standard deviation of the Gaussian function, respectively, which are preset according to the specific distribution characteristics of the health parameters;

[0139] S212. Design a neural network layer for each fuzzy set, receive the fuzzified input, make preliminary weight adjustments, and update the weights using the gradient descent method:

[0140]

[0141] Where η is the learning rate, E is the prediction error, and the weight w corresponds to the contribution of each health parameter in the fuzzy set;

[0142] S213, integrate all individual neural network layer outputs, merge information through the fusion layer, and fuse the outputs using a weighted summation method to form a unified output:

[0143]

[0144] Among them, output x is the output from the neural network layer corresponding to the health parameter x, w x is the corresponding weight, reflecting the importance of each health parameter in the final decision.

[0145] S3. Generate a nursing plan based on the in-depth analysis results of step S2, including treatment suggestions, lifestyle adjustment guidance and health monitoring plan;

[0146] In this implementation, S3 includes the following steps:

[0147] S31, receiving the output result of step S23, reflecting the patient's current health status and potential care needs;

[0148] S32. According to the specific value of V, a preset logic determination rule set is applied to determine the patient's treatment recommendations, lifestyle adjustment guidance, and health monitoring plan. The logic determination rule set R is implemented in the form of a multi-condition decision table, as follows:

[0149]

[0150] Among them, v1 and v2 are preset thresholds used to distinguish different levels of care needs;

[0151] S33. Generate specific nursing plans for each level of nursing needs:

[0152] The light care program includes basic health monitoring and regular assessment of lifestyle habits;

[0153] The moderate care program adds customized nutrition advice and moderate physical activity guidance;

[0154] Intensive care options include intensive health monitoring, medication management, and rehabilitation programs;

[0155] S34. Set corresponding monitoring plan parameters for each level of care plan. Parameters P include monitoring frequency, key health indicators and feedback mechanism:

[0156] P 轻度 =(F 月 , {HR, BP}, automatic feedback system);

[0157] P 中度 =(F 周 ,{HR, BP, SpO2, AL},Telemedicine support;

[0158] P 重度 =(F 日 , {HR, BP, SpO2, SQ, TR}, expert team intervention);

[0159] Among them, F represents the monitoring frequency, which is F per month. 月 、Weekly F 周 or daily F 日 ;

[0160] S35. The generated care plan and monitoring plan are automatically uploaded to the patient's electronic health record, and updated and notified in real time through the patient's mobile health application.

[0161] S4. Use a dynamic feedback mechanism to continuously monitor the patient's health status and the implementation of the nursing plan, collect and analyze changes in physiological parameters, patient self-feedback, and changes in environmental factors in the comprehensive health data set;

[0162] In this implementation, S4 includes the following steps:

[0163] S41. A real-time data monitoring system is set up to continuously collect the patient's physiological parameters, self-feedback and environmental factors. The real-time data monitoring system uploads the patient's heart rate HR, blood pressure BP, blood oxygen saturation SpO2, and other feedback and environmental data collected through the patient's mobile device to the cloud database in real time;

[0164] S42. Regularly evaluate the comprehensive health data set HDS and calculate the change Δ of each monitoring indicator:

[0165] ΔX=X current -X previous ;

[0166] Among them, X represents any monitoring indicator, X current is the current monitoring value, X previous is the value of the last record;

[0167] S43. Analyze the collected data using the data analysis module to identify any significant health trends or anomalies. The module uses preset health parameter thresholds for anomaly detection:

[0168] If ΔHR>HR threshold orΔBP>BP threshold then trigger alert;

[0169] Among them, HR threshold and BP threshold These are preset thresholds for heart rate and blood pressure that trigger health alerts.

[0170] S5. Adaptively adjust the nursing plan based on the data collected in step S4, and adapt to changes in the patient's condition by updating the nursing plan;

[0171] In this implementation, S5 includes the following steps:

[0172] S51. Systematic assessment of the patient's current health status and changes in environmental factors, including analysis of changes in monitored physiological parameters and environmental parameters such as air quality and temperature;

[0173] S52. Use the decision support system to adaptively adjust the nursing plan. The system uses predefined adjustment rules to update the nursing plan according to changes in health status. The rules determine whether adjustment is needed based on the following logical expression:

[0174] Adjustment Required=(|ΔX|>T X )for any X∈{HR, BP, SpO2, Env};

[0175] Where ΔX is the change in a given health or environmental parameter, T X is the threshold of the parameter, which is used to determine whether the change is significant;

[0176] S53. Automatically update the patient's care plan based on the output of the decision support system, including adjusting drug dosage, treatment frequency, lifestyle recommendations, and environmental control measures;

[0177] S54, realize environmental factor adjustment, automatically adjust the living habit guidance and health monitoring plan according to the changes of the patient's living environment:

[0178] When the environmental monitoring system detects a decrease in air quality, it automatically increases the use guidelines and frequency of indoor air purifiers;

[0179] Air quality has deteriorated, increasing the guidelines and frequency of use of indoor air purifiers;

[0180] Temperature fluctuations: When the temperature is too high or too low, the temperature control device settings at home will be automatically adjusted, and wearable suggestions will be provided;

[0181] Humidity changes: adjust the use of dehumidifiers or humidifiers based on high or low humidity, and remind patients about respiratory care in high or low humidity environments.

[0182] S6. Through telemedicine technology, medical professionals can remotely monitor the patient's condition and provide feedback and advice to the patient.

[0183] The present invention can analyze the comprehensive health data set (including physiological data, environmental factors and personal feedback) of lung cancer patients at home in real time by using an adaptive neuro-fuzzy inference system. The analysis is not only based on static health status, but is continuously monitored and updated according to the dynamic changes of the patient's condition to generate a care plan.

[0184] With the help of real-time data monitoring and adaptive adjustment mechanism, the present invention can effectively respond to any changes in the patient's condition. When significant changes in key health indicators (such as heart rate, blood pressure) or environmental factors (such as air quality, temperature) are monitored, the care plan can be automatically updated to ensure that the treatment recommendations are always consistent with the patient's latest health status.

[0185] The present invention focuses on the impact of environmental factors on the health of lung cancer patients and incorporates them into health data analysis and care plan adjustment, including monitoring air quality, temperature and humidity, etc., and automatically adjusting care recommendations based on these parameters, which helps to create a living environment that is more suitable for the health of patients.

[0186] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for generating nursing plans for lung cancer patients at home based on adaptive neural fuzzy reasoning, characterized in that: The following steps are involved: S1. Collect heart rate, blood pressure, blood oxygen saturation, sleep quality, activity level, eating habits and adverse reactions data of chemotherapy, immunotherapy and targeted therapy during treatment of lung cancer patients at home to build a comprehensive health data set; S2. performing an in-depth analysis of the comprehensive health data set by using an adaptive neural-fuzzy inference model, wherein the adaptive neural-fuzzy inference model combines the learning ability of a neural network with the reasoning ability of fuzzy logic to process health data containing uncertainty and ambiguity; S3. Generate a nursing plan based on the in-depth analysis results of step S2, including treatment suggestions, lifestyle adjustment guidance and health monitoring plan; S4. Use a dynamic feedback mechanism to continuously monitor the patient's health status and the implementation of the nursing plan, collect and analyze changes in physiological parameters, patient self-feedback, and changes in environmental factors in the comprehensive health data set; S5. Adaptively adjust the nursing plan based on the data collected in step S4, and adapt to changes in the patient's condition by updating the nursing plan; S6. Through telemedicine technology, medical professionals can remotely monitor the patient's condition and provide feedback and advice to the patient; The S2 comprises the following steps: S21. The adaptive neural fuzzy inference system is used to process the comprehensive health data set HDS. The adaptive neural fuzzy inference system combines the adaptive learning ability of the neural network and the reasoning ability of fuzzy logic to process the uncertainty and ambiguity in the data. The specific steps include: Use an independent fuzzification function fuzzify for each health parameter to generate the corresponding fuzzy set; The fuzzified health parameter set is used as input to construct the input layer of the fuzzy inference system; S22. The neural network component in the adaptive neural fuzzy reasoning system learns and trains the input fuzzy set, automatically extracts fuzzy rules, and adjusts the fuzzy logic rules and neural network weights to optimize the generation of nursing plans: Among them, Δw x is the weight update amount, corresponding to the health parameter x, μ x (x j ) is the fuzzy membership of the health parameter x in the jth sample, is the derivative of the activation function, λ is the regularization coefficient, used for L1 regularization of the weights, n is the number of samples, g(net j ) is the activation function of the neural network, net j is the net input to the neuron, calculated as: net j =∑ x∈X In x ·µ x (x j ); S23. Use the back propagation algorithm to train the neural network and optimize the weights of the fuzzy logic combination. For the error of each output node, apply the error gradient calculation formula: Among them, output k is the network output of the kth training sample, target k is the target output of the kth training sample, sigmoid'(net k ) is the derivative of the activation function, which calculates the net input at the kth sample, input ij is the i-th input connected to the j-th neuron, and N is the total number of training samples; S24, through steps S21-S23, the adaptive neuro-fuzzy inference model is dynamically adjusted according to the actual patient data to provide a nursing plan to cope with changes in the patient's health status; The S21 comprises the following steps: S211. For each health parameter x, including heart rate, blood pressure, blood oxygen saturation, sleep quality, activity level, eating habits, and adverse reaction data on chemotherapy, immunotherapy, and targeted therapy during treatment, a fuzzy set is generated through a predefined fuzzy function fuzzify(x). Each fuzzy function maps a real number input to the membership of a fuzzy set and uses a Gaussian function for fuzzification: Among them, μ x is the element x i The degree of membership to the fuzzy set x, c x and σ x are the center and standard deviation of the Gaussian function, respectively, which are preset according to the specific distribution characteristics of the health parameters; S212. Design a neural network layer for each fuzzy set, receive the fuzzified input, make preliminary weight adjustments, and update the weights using the gradient descent method: Where η is the learning rate, E is the prediction error, and the weight w corresponds to the contribution of each health parameter in the fuzzy set; S213, integrate all individual neural network layer outputs, merge information through the fusion layer, and fuse the outputs using a weighted summation method to form a unified output: Among them, output x is the output from the neural network layer corresponding to the health parameter x, w x is the corresponding weight, reflecting the importance of each health parameter in the final decision.

2. According to claim 1, a method for generating nursing plans for lung cancer patients at home based on adaptive neural fuzzy reasoning is characterized in that: The S1 comprises the following steps: S11. Collect physiological parameters of lung cancer patients at home, continuously collect heart rate HR data through ECG monitoring equipment, regularly record blood pressure BP data using an automatic sphygmomanometer, and monitor blood oxygen saturation SpO2 in real time through a pulse oximeter. When collecting data, each parameter is recorded at a frequency of one minute; S12. Automatically collect and analyze the patient's sleep quality data SQ and activity data AL, wherein the sleep quality data is monitored by a sleep monitoring device throughout the sleep cycle to evaluate the patient's sleep stage, duration, and interruptions: SQ=α×REM+β×NREM+γ×AWAKE; Among them, REM, NREM and AWAKE represent the duration ratio of rapid eye movement sleep, non-rapid eye movement sleep and wakefulness, respectively, and α, β and γ are weight coefficients, which are adjusted to optimize the sleep quality score; Activity data AL records the patient's steps, exercise type, and duration by syncing with wearable health trackers; S13. Record the patient's eating habits DH. Collect the type, amount and time of daily diet through the interaction between the patient and the nutrition management application. At the same time, analyze the diet photos uploaded by the patient through image recognition technology to automatically identify food categories and estimate intake: Among them, C j represents the intake of the jth food, ω j is the nutritional weight of the corresponding food, M is the total number of food categories; S14. Systematically collect adverse reaction data TR of patients during treatment, including adverse reactions of chemotherapy, immunotherapy, and targeted therapy. Automatically update the data through the electronic adverse reaction record form filled out by the patient, and analyze it in combination with the changes in the patient's physiological parameters to identify the correlation between possible adverse reactions and specific treatments; S15, comprehensively process and standardize the collected multi-source health data, including heart rate HR, blood pressure BP, blood oxygen saturation SpO2, sleep quality data SQ, activity data AL, eating habits DH and adverse reaction data TR, to construct a comprehensive health data set HDS: HDS={HR,BP,SpO2,SQ,AL,DH,TR}.

3. The method for generating nursing plans for lung cancer patients at home based on adaptive neural fuzzy reasoning according to claim 2, characterized in that: The S14 comprises the following steps: S141. Use the electronic health record system to automatically record the time series data of the heart rate HR, blood pressure BP, and blood oxygen saturation SpO2 of patients after chemotherapy, immunotherapy, and targeted therapy. According to the relationship between the changes in physiological parameters and the treatment time point, calculate the physiological response index R after each treatment. Use the time-weighted average method and add an exponential decay factor to make the physiological response index R sensitive to the recent physiological changes after treatment: Where T is the length of time from the start of monitoring after treatment, α, β and γ are the weights of different physiological parameters, reflecting their sensitivity in adverse reaction detection, and Z is the normalization constant to keep R within a reasonable range: S142. Automatically update data through the electronic adverse reaction record form filled out by the patient, recording the frequency and intensity of adverse reactions including nausea, vomiting, hair loss, and fatigue; S143. Combined with the physiological parameter response index R and the patient's self-reported adverse reaction data, a multivariate logistic regression model was used for analysis. Interaction terms and nonlinear terms were added to identify and predict the correlation between specific treatments and specific adverse reactions, and the patient's adverse reaction risk was assessed. The formula is as follows: Where P(y=1|x) is the probability of a specific adverse reaction occurring under a given treatment and physiological response index, β0, β j , β jk , γ j are model parameters, x j are predictor variables including treatment type, physiological response index R, and other related health indicators, x j x k Represents variable x j and x k The interaction term is used to model the interaction between two treatments or physiological responses. is the variable x j The quadratic term of is used to capture its nonlinear effect on the probability.

4. The method for generating nursing plans for lung cancer patients at home based on adaptive neural fuzzy reasoning according to claim 3, characterized in that: The S3 comprises the following steps: S31, receiving the output result of step S23, reflecting the patient's current health status and potential care needs; S32. According to the specific value of V, a preset logic determination rule set is applied to determine the patient's treatment recommendations, lifestyle adjustment guidance, and health monitoring plan. The logic determination rule set R is implemented in the form of a multi-condition decision table, as follows: Among them, v1 and v2 are preset thresholds used to distinguish different levels of care needs; S33. Generate specific nursing plans for each level of nursing needs: The light care program includes basic health monitoring and regular assessment of lifestyle habits; The moderate care program adds customized nutrition advice and moderate physical activity guidance; Intensive care options include intensive health monitoring, medication management, and rehabilitation programs; S34. Set corresponding monitoring plan parameters for each level of care plan. Parameters P include monitoring frequency, key health indicators and feedback mechanism: P 轻度 =(F 月 ,{HR,BP},automatic feedback system); P 中度 =(F 周 ,{HR,BP,SpO2,AL},telemedicine support); P 重度 =(F 日 ,{HR,BP,SpO2,SQ,TR},expert team intervention); Among them, F represents the monitoring frequency, which is F per month. 月 、Weekly F 周 or daily F 日 ; S35. The generated care plan and monitoring plan are automatically uploaded to the patient's electronic health record, and updated and notified in real time through the patient's mobile health application.

5. The method for generating nursing plans for lung cancer patients at home based on adaptive neural fuzzy reasoning according to claim 4, characterized in that: The S4 comprises the following steps: S41. A real-time data monitoring system is set up to continuously collect the patient's physiological parameters, self-feedback and environmental factors. The real-time data monitoring system uploads the patient's heart rate HR, blood pressure BP, blood oxygen saturation SpO2, and other feedback and environmental data collected through the patient's mobile device to the cloud database in real time; S42. Regularly evaluate the comprehensive health data set HDS and calculate the change Δ of each monitoring indicator: ΔX=X current -X previous ; Among them, X represents any monitoring indicator, X current is the current monitoring value, X previous is the value of the last record; S43. Analyze the collected data using the data analysis module to identify any significant health trends or anomalies. The module uses preset health parameter thresholds for anomaly detection: IfΔHR>HR threshold orΔBP>BP threshold thentriggeralert; Among them, HR threshold and BP threshold These are preset thresholds for heart rate and blood pressure that trigger health alerts.

6. The method for generating nursing plans for lung cancer patients at home based on adaptive neural fuzzy reasoning according to claim 5, characterized in that: The S5 comprises the following steps: S51. Systematic assessment of the patient's current health status and changes in environmental factors, including analysis of monitored physiological and environmental parameters, including changes in air quality and temperature; S52. Use the decision support system to adaptively adjust the nursing plan. The system uses predefined adjustment rules to update the nursing plan according to changes in health status. The rules determine whether adjustment is needed based on the following logical expression: AdjustmentRequired=(|ΔX|>T X )for anyX∈{HR,BP,SpO2,Env}; Where ΔX is the change in a given health or environmental parameter, T X is the threshold of the parameter, which is used to determine whether the change is significant; S53. Automatically update the patient's care plan based on the output of the decision support system, including adjusting drug dosage, treatment frequency, lifestyle recommendations, and environmental control measures; S54, realize environmental factor adjustment, automatically adjust the living habit guidance and health monitoring plan according to the changes of the patient's living environment: When the environmental monitoring system detects a decrease in air quality, it automatically increases the use guidelines and frequency of indoor air purifiers; Air quality has deteriorated, increasing the guidelines and frequency of use of indoor air purifiers; Temperature fluctuations: When the temperature is too high or too low, the temperature control device settings at home will be automatically adjusted, and wearable suggestions will be provided; Humidity changes: adjust the use of dehumidifiers or humidifiers based on high or low humidity, and remind patients about respiratory care in high or low humidity environments.

Images