Low-calcium nursing system for renal bone diseases of hemodialysis patients and operation method
By establishing a digital twin model and a personalized nursing system, the challenge of low calcium care for hemodialysis patients with renal osteodystrophy has been solved, improving the accuracy of nursing care and patient compliance, reducing the incidence of hypocalcemia, and enhancing treatment effectiveness and quality of life.
Patent Information
- Application Number
- CN202410997427.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-23
- Publication Date
- 2024-11-19
AI Technical Summary
How to effectively manage low calcium levels in hemodialysis patients with renal osteodystrophy, especially after using denosumab, in order to reduce adverse reactions and complications.
A low-calcium nursing system for hemodialysis patients with renal osteodystrophy was designed. By collecting patient data, establishing a digital twin model, and combining the data with specific patient parameters for data processing and analysis, personalized nursing advice and psychological support are provided, including health education, virtual companionship and other services.
It improved the precision of nursing care and patient compliance, reduced the incidence of hypocalcemia, and enhanced the treatment effect and quality of life of patients.
Smart Images

Figure CN118983049A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to software technology, and in particular to a low calcium nursing system and an operating method for renal osteopathy in hemodialysis patients. Background Art
[0002] Renal osteodystrophy is a serious complication of patients with chronic renal failure, and its main feature is multi-system calcification outside the skeleton. Renal osteodystrophy is divided into β2-microglobulin amyloid osteoarthropathy, mixed bone disease, low-turnover bone disease, and high-turnover bone disease according to different bone turnover rates. Metabolic disorders of the body are the cause of renal osteodystrophy. Its pathogenesis is due to metabolic disorders of vitamin D, phosphorus, calcium, etc., acid-base imbalance, hyperparathyroidism, and renal osteodystrophy. The clinical manifestations of renal osteodystrophy mainly include skin pain, bone deformation, skin itching, fractures, bone pain, etc.
[0003] Studies have found that the use of denosumab injection in the treatment of diabetic osteoporotic vertebral fractures after surgery can effectively improve its clinical treatment effect, accelerate bone healing time, increase serum bone density, thereby increasing bone mass and bone strength, and reducing the incidence of re-fractures. Compared with traditional anti-osteoporosis drugs bisphosphonates, denosumab has the advantages of not causing flu-like symptoms, no need to adjust the dose for patients with renal impairment, and good gastrointestinal tolerance.
[0004] Hemodialysis patients may suffer from hypocalcemia due to vitamin D deficiency and drugs such as cilaxet. Hypocalcemia is also a major complication after total parathyroidectomy. Studies have reported that denosumab has adverse reactions such as pain, fractures, and hypocalcemia in clinical use. The incidence of hypocalcemia in the treatment of osteoporosis in postmenopausal women with denosumab is 2% to 20%, and it is as high as 10% to 50% in patients with bone metastases who receive a single dose of 120 mg. 82.93% of hypocalcemia occurs within 1 month of medication.
[0005] Therefore, how to provide good low calcium care for hemodialysis patients with renal osteopathy, especially after using denosumab, is a technical problem that needs to be solved urgently. Summary of the invention
[0006] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is to provide a low calcium care system and operation method for renal osteopathy in hemodialysis patients, which can collect patient data, infer patient habits and disease progression, so as to provide reference or suggestions for subsequent care and patient precautions.
[0007] To achieve the above object, the present invention provides a low calcium care system for renal osteodystrophy in hemodialysis patients, comprising:
[0008] The patient side is used to collect patient information, enable communication between the patient and the doctor, and obtain information from the server;
[0009] The medical terminal is used for medical staff to upload patient data and obtain the data uploaded by the patient terminal and the corresponding analysis results from the server;
[0010] The server is used to parse the data uploaded by the patient and doctor, and then store it in the ID directory identified by the ID data module, and process the corresponding data through the digital twin model and nursing module to output the results;
[0011] Network, used to connect medical and nursing terminals, servers, and patients;
[0012] ID data module, used to identify the ID to which the data belongs and perform data processing;
[0013] The digital twin model collects a large amount of patient data, disease progression, and nursing plans to build a software model. Then, based on the specific patient data, adjustments made by the medical staff, and system identification, the software model is further optimized to become the digital twin model of the patient. The digital twin model is then used to predict disease progression and nursing plans.
[0014] The nursing module is used to provide health education, social interaction, and virtual companion services to patients based on the data collected from patients by medical staff.
[0015] As a further improvement of the present invention, the software model includes:
[0016] Data collection, collecting patients’ physiological parameters and pathological data;
[0017] Data processing: clean and assign values to the data. Ratio the measured value to the standard parameter value and retain two decimal places to obtain the assigned value. The measured parameter that is less than the minimum value of the standard parameter interval is ratioed to the minimum value of the standard parameter interval. The measured parameter that is greater than the maximum value of the standard parameter interval is ratioed to the maximum value of the standard parameter interval. The value between the minimum and maximum values of the standard parameter interval is 1. Weight some selected parameters by multiplying their assigned values by 2 to obtain the final assigned value.
[0018] Data integration, which combines and fuses data from different data sources into a comprehensive and consistent patient data set;
[0019] Feature extraction, selecting features relevant to medical decision-making from the integrated patient dataset;
[0020] Modeling and training: using extracted features and medical expertise to build software models;
[0021] Model verification and evaluation: treatment cases are selected and input into the model to verify the accuracy and predictive ability of the model, while the model performance is evaluated and patients are continuously optimized.
[0022] As a further improvement of the present invention, the digital twin model is obtained by relying on a software model and combining selected parameter types; according to the selected parameter types, the model corresponding to this part of the parameter type is stripped from the software model and optimized to obtain the digital twin model.
[0023] As a further improvement of the present invention, the nursing module comprises:
[0024] The semantic recognition module is used to recognize the semantics and perform corresponding operations according to the recognized semantics;
[0025] The speech recognition module is used to recognize the input speech and convert it into text for semantic recognition or directly perform semantic recognition;
[0026] An image recognition module is used to recognize input pictures and photos. The image recognition module is used to recognize input photos of medicines and food, so as to record the patient's medication and diet, or provide guidance on the patient's medication and diet;
[0027] Movement recognition module, used to obtain the patient's movement status and movement amount;
[0028] The status monitoring module is used for medication monitoring, diet monitoring, and health monitoring. Medication monitoring is based on the patient's own feedback and the photos of medication taken to monitor whether the patient is taking the medication according to the doctor's orders; diet monitoring monitors the patient's diet category by actively inputting food information through the patient's end and recognizing food information through images, and then calculates the impact of food on the patient based on the corresponding digital twin model to guide the patient to change his eating habits and eat healthily; health monitoring monitors the patient's health status through medical data and wearable data;
[0029] The assistant module is used for health education, state adjustment, social adjustment, and habit inference. The patient can send a request to the server or the medical staff can actively push relevant popular knowledge to the patient through the server. The patient's mental state is judged through the information detected by the patient and the information actively input by the patient, and adjustment information is sent to the patient in a timely manner according to the corresponding state. The patient actively infers the patient's habits through the detected data, and then gives corresponding improvement or encouragement measures based on the patient's habits combined with the digital twin model.
[0030] Virtual companions, combined with existing artificial intelligence technology, can initiate communication and companionship when the patient actively requests it or the system detects that the patient may need it, thereby further realizing psychological care.
[0031] The present invention also discloses a method for operating a low calcium system for hemodialysis patients with renal osteodystrophy, comprising:
[0032] S100, Building a Digital Twin Model of Patients
[0033] S110. According to the current research on the nursing of low calcium in hemodialysis patients with renal osteopathy, manually calibrate or select the corresponding data types through artificial intelligence technology, and then assign weights to each data type;
[0034] S120. Separating the patient's digital twin model from the software model according to the selected data types; the selected data types in the digital twin model have been trained, and include the influence ratio of each data type on the treatment result and the weight of each data type;
[0035] S130, predict and calculate based on the patient's data, and then compare with the actual results to make corrections and optimizations;
[0036] S140, Health Belief Model
[0037] S141. Educate patients. Output educational information to patients based on their conditions to help them understand their conditions, matters that require attention, lifestyle habits that need to be improved, and how to maintain a good attitude and increase compliance.
[0038] S142, psychological support, assess the patient's mental state, and actively guide the patient to maintain a good attitude and confidence in the treatment effect;
[0039] S143, monitoring and feedback, monitoring the patient's compliance, psychological stress and the impact of the health belief model on the development of the patient's condition through patient follow-up and patient self-test feedback, and continuously improving the deficiencies;
[0040] S150, the digital twin model combines the input data to give treatment, nursing suggestions, and HBM suggestions;
[0041] S200, treatment and monitoring, output treatment suggestions, nursing suggestions, and patient suggestions based on the digital twin model, and then medical staff specify treatment methods, nursing methods, HBM care and suggestions for patients based on the suggestions; use the digital twin model to infer the patient's disease progression and treatment effect based on the treatment method, nursing method, patient's pathological data, and patient's trial data, and regularly monitor the patient's actual parameters;
[0042] S300, lifestyle and nutrition management, outputs recommendations on patients’ lifestyle and nutrition management based on the digital twin model to guide medical staff and patients to change their lifestyles and improve their nutrition according to the recommendations;
[0043] S400, Complication Prevention and Care, predicts possible complications through the digital twin model, and takes preventive and nursing measures; the actual complications of the patient are input into the system to verify the accuracy of the digital twin model;
[0044] S500, continuous monitoring and evaluation, continuously monitors the patient's data and compares it with the predictions of the digital twin model. If deviations are found, corrections are made to improve the prediction accuracy of the digital twin model.
[0045] As a further improvement of the present invention, in S110, the assignment process is as follows: if the data can be parameterized, the two decimal places after the ratio of the measured parameter to the standard parameter are used as the weight value, and then some parameters that need to be weighted are weighted to obtain the final assignment, the measured parameter that is less than the minimum value of the standard parameter interval is ratioed to the minimum value of the standard parameter interval, the measured parameter that is greater than the maximum value of the standard parameter interval is ratioed to the maximum value of the standard parameter interval, and the value between the minimum and maximum values of the standard parameter interval is 1; and the HBM factor is a relatively subjective parameter, and the total score is assigned and then each factor is equally divided to assign the value.
[0046] As a further improvement of the present invention, in S120, the Cox proportional hazard model is used to verify whether the influence of each parameter on the result is significant through hypothesis testing, and the parameters with significant influence on the result and the selected parameters are calculated through Logistic regression, and the OR value is used as the influence ratio.
[0047] As a further improvement of the present invention, S200 also includes:
[0048] S210. Methods of monitoring dialysis treatment;
[0049] S220, calcium and vitamin D supplementation, adjust the dose and frequency of oral calcium according to the patient's blood calcium level; regularly monitor blood calcium concentration, adjust the dose according to the results, and correct the digital twin model;
[0050] S230, iPTH monitoring;
[0051] S240, avoid drug interactions;
[0052] S250. Assess the degree of renal function impairment by estimating the glomerular filtration rate (eGFR).
[0053] As a further improvement of the present invention, S300 further includes:
[0054] S310, Nutrition and Diet Management, monitor diet, limit phosphorus intake, increase calcium intake, and maintain dietary diversity;
[0055] S320. Change your lifestyle, quit smoking and drinking, and get enough sleep;
[0056] S330: Exercise and training: monitor the patient's exercise volume and duration to determine whether the exercise meets the requirements.
[0057] As a further improvement of the present invention, S500 also includes:
[0058] S510, regular review: according to the review cycle set by the medical and nursing end, remind the patient to have a regular review, and upload the review information to compare with the previous information, verify the prediction of the digital twin model, further predict the condition, provide corresponding nursing guidance, and provide corresponding health education information to the patient;
[0059] S520, Assessment:
[0060] S521. Evaluation of predictions and extrapolations of digital twin models:
[0061] 1) Input the patient monitoring information and then compare it with the information predicted by the digital twin model to find inaccuracies in the predicted information and then make improvements;
[0062] 2) The digital twin model multiplies the value of each data type input by the proportion of the data's impact on the treatment outcome, and then takes the percentage; for the prediction results corresponding to multiple data, the percentages of each data are added together to obtain the comprehensive percentage; if the percentage is greater than or equal to 100%, it is predicted that the situation predicted by the data will definitely occur; if it is greater than or equal to 50%, it means that the situation predicted by the data has a high probability of occurring; if it is less than 50%, it means that the situation predicted by the data has a low probability of occurring;
[0063] After the follow-up review, the patient's actual situation is input, and then the system compares it. If the predicted situation with a percentage greater than or equal to 100% does not occur, it means that the prediction is wrong. At this time, the system reminds the medical staff or maintenance personnel to make corrections; if a low-probability situation occurs multiple times, it means that the situation should be a high-probability event or a situation that should occur. At this time, the system also reminds the medical staff or maintenance personnel to make corrections;
[0064] S522. Evaluation of disease changes and nursing plans. Through the digital twin model, historical parameters, and revised parameters, the detected disease changes and nursing plans are output; cases where the treatment of the disease is not as effective as expected are pushed to medical staff, who will optimize or change the nursing plan and monitor the results of the implementation of the new plan. If the results are positive, the modified plan will be used as the preferred nursing plan and updated in the digital twin model at the same time.
[0065] The beneficial effects of the present invention are:
[0066] The nursing system of the present invention uses a network to connect the server, the medical end, and the patient end. After that, the ID data module is used to identify, process, and store the collected data in the corresponding patient ID directory for the convenience of subsequent system calls. At the same time, the digital twin model is used to combine the patient's parameters to output corresponding predictions, inferences, and medical and nursing suggestions to provide guidance for the treatment and nursing of medical staff and the cooperation of patients; at the same time, the patient's data is collected through the nursing module, and psychological care and education are provided to the patient to further increase the nursing effect and patient compliance. The operation method of the nursing system is separated and optimized from the software model according to the data type selected for the patient, so as to obtain the digital twin model, and the selected data is assigned and weighted, and the influence ratio of each parameter on the result is obtained. Subsequently, each parameter, the combination of each parameter, and the influence ratio of each corresponding result are used to calculate the percentage of the result, so as to output the predicted result and the probability of occurrence. Then, optimization is performed according to the actual input situation to continuously improve the prediction accuracy, so as to provide more accurate predictions, inferences, etc. for subsequent nursing, HBM suggestions, etc. BRIEF DESCRIPTION OF THE DRAWINGS
[0067] Figure 1 It is a schematic diagram of the composition of the low calcium care system;
[0068] Figure 2 It is a schematic diagram of the composition of the digital twin model;
[0069] Figure 3 It is a schematic diagram of the composition of the nursing module;
[0070] Figure 4 is a schematic diagram of the method of operation of the low calcium system. DETAILED DESCRIPTION
[0071] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.
[0072] See also Figure 1 The low calcium care system of this embodiment includes:
[0073] The patient end is used to collect patient information, realize communication between the patient and the doctor, and obtain information from the server. The patient end of this embodiment includes a wearable end, a smart phone, etc. The wearable end can be a bracelet, through which the patient's heart rate, respiratory rate, electrocardiogram, blood pressure, blood sugar, motion data, images, etc. are collected. The smart phone and the wearable end are bound to the patient's ID, so that the patient can obtain corresponding popular science materials, precautions and other materials through the ID to complete the education of the patient.
[0074] The medical terminal is used for medical staff to upload patient data, obtain the data uploaded by the patient terminal and the corresponding analysis results from the server. The medical terminal can be a computer, tablet computer, smart phone, etc.
[0075] The server is used to parse the data uploaded by the patient and doctor, and then store it in the ID directory identified by the ID data module, and process the corresponding data through the digital twin model and nursing module to output the results.
[0076] The network is used to connect the medical and nursing end, server, and patient end.
[0077] The ID data module is used to identify the ID to which the data belongs and to perform data processing, such as data cleaning and filtering. The ID data module is a software that can be installed on the medical end, the patient end, and the server. When used, it first identifies the ID corresponding to the data, then processes the data, and finally stores the data in the preset categories under the ID directory according to the data classification. For example, for the blood calcium concentration data, after the medical end uploads it, the system matches it in the server according to the ID corresponding to the blood calcium data until the directory storing the data of the ID is found, and then parses and classifies the data into blood calcium concentration, so that the blood calcium concentration is stored in the blood calcium concentration data of the ID.
[0078] The digital twin model collects a large amount of patient data, disease progression, nursing plans, etc. to build a software model, and then further optimizes the software model into a digital twin model of the patient based on the specific patient data, adjustments made by the medical and nursing side, and system identification. The digital twin model is then used to predict disease progression, nursing plans, etc.
[0079] The nursing module is used to provide health education, social interaction, virtual companionship and other services to patients based on the medical and nursing end and the data collected from patients. The digital twin model and nursing module are both software that can be installed on the patient end, medical and nursing end, server, etc. to provide corresponding functions.
[0080] The internal system is the intranet or internal management system of a medical institution. Data input through the intranet can be directly assigned to the digital twin model, nursing module, etc., which greatly facilitates use.
[0081] See also Figure 2 , the software model includes:
[0082] Data collection mainly involves collecting patients' physiological parameters such as heart rate, respiratory rate, electrocardiogram, blood pressure, height, weight, etc., as well as pathological data collection such as medical history, blood calcium concentration, iPTH monitoring data, bone density, etc.
[0083] Data processing: clean the data to filter out some interference data, useless data, etc.; assign values: compare the test data with the standard data, retaining two decimal places.
[0084] Compare the data with the healthy standard data and take the difference, and assign a value according to the size of the difference to facilitate subsequent evaluation. For example, the reference range of normal blood calcium is 2.25-2.75mmol / L. If the data is within this range, it is judged as normal and assigned a value of 1; if it is less than this range, it is judged as hypocalcemia, and the assigned value is the ratio of the data divided by the minimum value of the reference range and the two decimal places; if it exceeds this range, it is judged as hypercalcemia, and the assigned value is the ratio of the data divided by the maximum value of the reference range and the two decimal places. The size of the ratio can be used to roughly judge the blood calcium situation in the future, which is also convenient for subsequent calculations. For other parameters, this method can be used for assignment. Then weight each parameter, such as blood calcium, blood phosphorus, parathyroid hormone (iPTH), etc., multiply it by 2 on the basis of its assignment to obtain the final assignment.
[0085] Data integration integrates and merges data from different data sources to form a comprehensive and consistent patient data set. Data warehouses, data lakes and other technical means are used to achieve unified management and query of different data formats and structures.
[0086] Feature extraction, select features related to medical decision-making from the integrated patient data set. Statistical methods, machine learning algorithms, etc. can be used to find features related to the patient's condition and treatment effect through data mining and analysis. Of course, it can also be determined by manual selection, such as blood calcium, blood phosphorus, parathyroid hormone (iPTH) in the low calcium care of renal osteodystrophy in hemodialysis patients, and quantitative factors affecting patients in the health belief model.
[0087] Modeling and training: Use the extracted features and medical expertise to build a software model. You can use traditional statistical analysis methods or machine learning algorithms such as deep learning and neural networks. Then train and tune the model to improve its accuracy and predictive ability. Through continuous iteration and optimization, the model will be closer to the actual situation of the patient.
[0088] Model verification and evaluation: treatment cases are selected and input into the model to verify the accuracy and predictive ability of the model, while the model performance is evaluated and patients are continuously optimized.
[0089] The establishment of software models is a relatively common technology, which has been widely used in the medical industry and can be used to predict and optimize the treatment and surgery of patients. Therefore, this embodiment directly adopts the existing technology. At the same time, this embodiment is mainly aimed at the care of low calcium in hemodialysis patients with renal osteodystrophy, so the types of data required are not many, and the current research is relatively mature, so it is completely possible to use manual modeling to model, that is, manually set the required data types, the corresponding weights of each data and the impact on the results, etc.
[0090] The digital twin model is an optimized special case of the software model, which mainly relies on the software model obtained by big data and combines the patient's special situation to obtain a digital twin model of a patient. For example, according to the selection of corresponding data on the medical side or the use of the data uploaded by the medical side and / or the patient side corresponding to the software model, the model with this part of the data in the software model is selected as the digital twin model. Because the parameter types corresponding to the software model are relatively comprehensive, and the digital twin model of the patient often does not require so many parameters, it is sufficient to directly select the required parameter types in the software model to form a digital twin model. At the same time, this model has also been trained with big data and has corresponding prediction and inference functions, which can meet the needs of use.
[0091] See also Figure 3 , the nursing module comprises:
[0092] The semantic recognition module is used to recognize semantics and perform corresponding operations based on the recognized semantics. For example, the semantics of the relevant educational materials that you want to obtain are input in the recognized text, and then the corresponding educational materials are output according to the requirements. This is already quite common in existing artificial intelligence technologies, and existing semantic recognition technologies can be directly used.
[0093] The speech recognition module is used to recognize the input speech and convert it into text for semantic recognition or directly perform semantic recognition. This is very common in the speech interaction technology in the prior art, so the prior art can be selected.
[0094] The image recognition module is used to recognize input pictures and photos. In this embodiment, the image recognition module can be used to recognize input drug photos, food photos, etc., so as to record the patient's medication and diet, or provide guidance on the patient's medication and diet. For example, before taking the medicine, the medicine is placed and photographed and uploaded to the server. The server compares the shape and color characteristics of the medicine with the image of the medicine in the prescription pre-stored on the server. If an error is found, the patient is reminded in time and the medical care end is notified, such as the quantity error, the medicine does not correspond to the medicine in the prescription, etc.; if no error is found, the medication information is recorded to monitor the patient's medication. After uploading the food photo to identify the food, the digital twin model of the patient can be used to predict whether the food will affect the patient's condition. If so, the patient is warned and the medical staff is prompted. For example, for patients with hypocalcemia, high-salt, high-carbohydrate, and high-purine foods are generally not allowed. Once the above foods are identified in the food photos, the patient is reminded to eat with caution, and the medical care end is notified and recorded on the server, so that the medical care end can intervene in the care in time and the user can improve his eating habits.
[0095] The motion recognition module is a hardware module set in the patient's wearable terminal and the corresponding software, such as a gyroscope, an accelerometer, etc. This is a prior art, and the motion detection technology of the Xiaotiancai phone watch can be used specifically. The motion recognition module is used to obtain the patient's motion status and amount of exercise, and the motion status includes no exercise, walking, running, limb exercise (such as yoga), etc. The amount of exercise is mainly calculated by detecting data such as exercise duration, number of steps, body temperature changes, breathing rate and heart rate. It is a commonly used technology to calculate the amount of exercise based on the number of steps and exercise duration, such as the calorie consumption calculation in the walking navigation of Gaode Map. As for the change in body temperature, this is mainly to judge the intensity of the exercise by detecting the change in body temperature over a period of time. Of course, the detection effect is more accurate when combined with the heart rate and breathing rate. For every degree Celsius increase in body temperature, the basal metabolic rate may increase by about 12%, from which the amount of exercise can be roughly calculated. In addition, using heart rate and breathing rate, the formula for calculating exercise volume is: exercise volume = exercise intensity × duration. Exercise intensity can be monitored by heart rate. The heart rate of moderate-intensity aerobic exercise is 60%-70% of the maximum heart rate. At the same time, the breathing rate can provide feedback on the intensity of exercise. This is also an existing technology. Specifically, the exercise volume assessment technology in existing sports bracelets can be used, such as the FirstBeat algorithm.
[0096] The status monitoring module is used for medication monitoring, diet monitoring, health monitoring, etc. Medication monitoring is mainly based on the patient's own feedback, taking pictures of medication, etc. to monitor whether the patient is taking the medication according to the doctor's orders. Dietary monitoring can monitor the patient's diet category, food intake, etc. through the patient's active input of food and image recognition of food, and then calculate the impact of food on the patient based on the corresponding digital twin model to guide the patient to change his eating habits and eat healthily. Health monitoring mainly monitors the patient's health status through medical data and wearable data, such as judging the progression of the disease based on blood calcium data and blood phosphorus data uploaded by the medical care end, and monitoring the current physical state through the wearable end's body temperature, heart rate, electrocardiogram, etc.
[0097] The assistant module is used for health education, status adjustment, social adjustment, and habit speculation. The patient can send a request to the server or the medical staff can actively push relevant popular knowledge to the patient through the server, such as explaining in detail the mechanism of action of renal osteopathy, denosumab, and the risk of hypocalcemia, emphasizing the importance of regular monitoring of blood calcium, and the serious consequences of abnormal blood calcium levels. Encourage patients to have confidence in overcoming the disease and realize that the disease can be effectively controlled through reasonable treatment and care. Emphasize that medical staff will pay attention to changes in the patient's condition throughout the process and provide necessary support and help.
[0098] The patient's mental state can be judged through the information detected by the patient side and the information actively input by the patient, such as using the wearable end (bracelet) to detect the patient's sleep quality, depression state, etc., and timely send adjustment information to the patient side according to the corresponding state, such as playing music for the patient, pushing small games, etc., and at the same time feedback the relevant information to the medical side, and the medical side can provide further psychological care. This can pay attention to the patient's mental state and provide psychological counseling and support in time. At the same time, encourage patients to maintain an optimistic attitude and actively face the disease and treatment. At the same time, certain social adjustments can also be made, such as establishing a patient circle based on the distance of the surrounding patient side, popularizing relevant knowledge through communication and exchanges between patients, and encouraging each other, holding online or offline activities, etc. to further increase patients' knowledge of the development of the disease, precautions, etc., and reduce the psychological burden of patients. Of course, the patient's social account can also be linked, and health education and psychological care for patients can be completed through communication between patients and communication between doctors and patients.
[0099] The patient side can actively infer the patient's habits, such as work and rest habits, medication habits, eating habits, etc., through the detected data, so as to give corresponding improvements or incentives based on the patient's habits combined with the digital twin model. For example, based on the patient's sleep time, if the sleep time is too late, the patient will be actively reminded to go to bed as early as possible and explain the corresponding benefits and disadvantages; or if it is detected that the patient likes to eat high-fat foods, dietary recommendations will be pushed to the patient and the corresponding benefits and disadvantages will be explained. In this way, the patient's habits are gradually changed to assist in the treatment of related diseases.
[0100] Virtual companions, combined with existing artificial intelligence technology, can initiate communication and companionship when patients actively request it or when the system detects that patients may need it, thereby further achieving psychological care. For example, when it is detected that a patient is in a state of anxiety or depression, the patient can actively initiate a conversation with the patient through the patient end to guide the patient.
[0101] During use, the server continuously receives information from the medical and patient ends, and inputs the software model and the patient's corresponding digital twin model for training for optimization and inference, and then delegates the optimized digital twin model to the medical and patient ends for control, analysis, and inference.
[0102] See also Figure 4 , based on the operating method of the above-mentioned low calcium care system, including:
[0103] S100, Building a Digital Twin Model of Patients
[0104] S110. According to the current research on the nursing of low calcium in hemodialysis patients with renal osteodystrophy, manually calibrate or select corresponding data types through artificial intelligence technology, such as blood calcium, blood phosphorus, parathyroid hormone (iPTH), bone density, medication information, medical history, basic patient information (height, weight, allergy information, pain tolerance, renal function, liver function, etc.), HBM factors, etc., and then assign weights to each data type. Among them, the parameterized information such as blood calcium, blood phosphorus, parathyroid hormone (iPTH), bone density, medication information, medical history, basic patient information, etc. can use the two decimal places of the ratio of the measured parameter to the standard parameter as the weight value, and then weight some parameters that need to be weighted to obtain the final assignment, the measured parameter less than the minimum value of the standard parameter interval is the ratio of the minimum value of the standard parameter interval, the maximum value of the standard parameter interval greater than the maximum value of the standard parameter interval is the ratio, and the value between the minimum value and the maximum value of the standard parameter interval is 1. The HBM factor is a relatively subjective parameter. This embodiment adopts the method of equally dividing each factor after assigning a total score. For example, the HBM factor is assigned 1 point, and then the score of each influencing factor selected is 1 divided by the number of selected factor items. For example, the number of selected factors is education, compliance, habit change, exercise, mental health status, social interaction, and emotional support, a total of 8 items, then each item is assigned 1 / 8 point, which is 0.13 points after rounding to two decimal places.
[0105] S120. Based on the selected data types, the digital twin model of the patient is separated from the software model; the data types selected in the digital twin model have been trained, and include the impact of each data type on the treatment outcome (impact ratio, the ratio of the data to the corresponding consequence, the larger the ratio, the more likely it is to cause the consequence, which can be set manually or calibrated using current statistical techniques and correlation analysis techniques), the weight of each data type, etc., so that predictions and calculations can be made through the patient's data.
[0106] In this embodiment, it is possible to test which parameters have a significant effect on the results by hypothesis testing. For example, the Cox proportional hazard model is used to verify whether the effect of each parameter on the results is significant by hypothesis testing. The parameters that have a significant effect on the results and the selected parameters are calculated by Logistic regression. The OR value (Odds Ratio) is used as the impact ratio. In Logistic regression, the OR value is an important indicator for measuring the degree of influence of different variables in the model on the results. By calculating the OR value, the specific influence of each parameter on the results can be understood.
[0107] In most statistical software (such as R, Stata, SAS, SPSS, etc.), when performing logistic regression, the coefficient (β) of each independent variable and its standard error, z value, p value and other statistics can be directly output. To obtain the OR, just take the exponent of the corresponding coefficient β (i.e., e β ), this value is the multiple of the change in the odds ratio of the dependent variable event occurring versus not occurring when the independent variable increases by one unit.
[0108] S130, predict and extrapolate the patient's data, and then compare it with the actual results to make corrections and optimizations. If the extrapolated results deviate from the patient's actual results, the influence ratio and weight of the corresponding parameters can be adjusted manually to continuously optimize the model and improve the prediction ability.
[0109] S140, Health Belief Model (HBM)
[0110] Studies have shown that establishing a health belief model (HBM) for patients can make a significant contribution to realizing their susceptibility to hypocalcemia in renal osteodystrophy, the threat of hypocalcemia in renal osteodystrophy, the benefits of adopting healthy behaviors, the obstacles to adopting healthy behaviors, the action cues to promote healthy behaviors, and the improvement of self-efficacy. It also has a very positive promoting significance for treatment and nursing.
[0111] S141. Educate patients
[0112] 1) Through doctor’s advice, health promotion or personal symptom experience, they realize that their calcium metabolism is abnormal due to kidney disease, which increases their risk of developing renal osteodystrophy. This perception of disease susceptibility will prompt individuals to pay more attention to their health status.
[0113] 2) Understand the threat of low calcium in renal osteodystrophy. Renal osteodystrophy is caused by hypocalcemia and 1.25-hydroxyvitamin D3 deficiency during the progression of chronic kidney disease, which leads to secondary hyperparathyroidism and severe bone damage. This includes symptoms such as bone pain, fractures, and bone deformation, which seriously affect the patient's quality of life. In addition, renal osteodystrophy may also cause extra-bone lesions such as anemia, nervous system damage, and cardiovascular disease, further increasing the patient's health risks. Clinical data show that patients with high parathyroid hormone have significantly increased total mortality or cardiovascular disease mortality, and parathyroid hormone levels are positively correlated with the degree of vascular calcification.
[0114] 3) Recognize that improving calcium intake, supplementing vitamin D, controlling the progression of kidney disease and other measures can effectively prevent or alleviate the symptoms of renal osteodystrophy and improve the quality of life. At the same time, these measures may also reduce the risk of extra-bone lesions such as cardiovascular disease and further protect the overall health of patients.
[0115] 4) Patients may face a variety of obstacles, such as drug side effects, dietary restrictions, financial burden, etc. These factors may make individuals resist taking healthy behaviors. However, through health education, doctor guidance and social support, individuals can be helped to identify and resolve these obstacles, thereby improving their willingness and ability to take healthy behaviors.
[0116] 5) Doctors, family members, friends or health promotion can all serve as action clues to promote individuals to adopt healthy behaviors. For example, doctors can prescribe and guide patients to use calcium and vitamin D supplements correctly; family members can monitor patients' diet and medication; health promotion can popularize relevant knowledge about renal osteodystrophy and encourage patients to take active treatment measures.
[0117] 6) By successfully implementing healthy behaviors and receiving positive feedback, an individual's self-efficacy will be improved. For example, after the patient adheres to calcium and vitamin D supplementation, the bone pain symptoms are relieved, which will enhance their confidence in continuing to adopt healthy behaviors. In addition, the encouragement and help of doctors and social support systems can also help improve the patient's self-efficacy and enable them to more actively cope with the challenges of renal osteodystrophy.
[0118] S142. Psychological support
[0119] 1) Understand and evaluate mental health status. The SCL-90 scale can be used to understand the patient's mental health level from ten aspects, and depression, anxiety and stress scales such as DASS-21 can be used to assess risk factors and provide corresponding suggestions.
[0120] 2) Psychological counseling and emotional support: Provide corresponding psychological counseling according to the different mental states of patients to reduce their psychological pressure. At the same time, provide emotional support through communication between doctors and nurses, communication between patients, and communication between family members and patients, so as to increase patients' determination and confidence in treatment.
[0121] 3) Establish a trusting relationship through communication between doctors and patients, between patients, and between family members and patients, so that patients can follow doctor's orders and relieve psychological stress.
[0122] 4) Dynamic-oriented treatment indirectly promotes the overall treatment effect of renal osteopathy by improving the patient's psychological state and treatment compliance. It helps patients better cope with the psychological pressure brought by the disease, thereby improving the overall treatment effect.
[0123] S143. Monitoring and feedback: monitor patient compliance, psychological stress, and the impact of the Health Belief Model (HBM) on the development of the patient's condition through patient follow-up and patient self-test feedback, and continuously improve deficiencies.
[0124] S150. The digital twin model combines the input data to give treatment and nursing suggestions, HBM suggestions, etc. HBM suggestions are suggestions for the digital twin model to establish a health belief model based on the HBM factor output, including suggestions for medical staff, patients, and patients' families.
[0125] S200, Treatment and Monitoring
[0126] S210. Monitor the method of dialysis treatment, such as hemodialysis or peritoneal dialysis, and monitor the dialysate calcium concentration and dialysis duration. Hemodialysis has a greater impact on calcium and phosphorus metabolism than peritoneal dialysis, so hemodialysis patients are more likely to develop hypocalcemia. Appropriate dialysis duration and a reasonable dialysis regimen (such as the use of low-calcium dialysate) can effectively improve the patient's calcium and phosphorus metabolism and reduce the occurrence of hypocalcemia. The number of dialysis sessions can be increased or the dialysis time can be extended to increase the clearance of phosphorus. You can also consider using more efficient dialysis methods such as high-flux dialysis or hemodiafiltration (HDF).
[0127] S220, calcium and vitamin D supplementation, adjust the dose and frequency of oral calcium according to the patient's blood calcium level. Monitor blood calcium concentration regularly, generally at least once a month, and adjust the dose according to the results. Monitor serum 25-hydroxyvitamin D (25(OH)D) levels, and start vitamin D supplementation when serum total 25-hydroxyvitamin D does not reach 50.8ng / mL. Calcium and vitamin D supplementation can be evaluated by monitoring medication use, and then corrected by regular checks of blood calcium and serum 25-hydroxyvitamin D (25(OH)D) levels.
[0128] S230, monitoring of parathyroid hormone (iPTH). The normal range of parathyroid hormone is usually between 12-88pg / mL. According to many studies, chronic renal failure patients on long-term maintenance hemodialysis will have excessive secretion of parathyroid hormone due to the stimulation of increased blood phosphorus and decreased blood calcium. This excessive secretion of parathyroid hormone can cause secondary hyperparathyroidism (SHPT), which in turn leads to the occurrence of renal osteodystrophy. Therefore, monitoring and controlling parathyroid hormone (iPTH) can effectively reduce the probability of hypocalcemia in hemodialysis patients with renal osteodystrophy.
[0129] S240. Avoid drug interactions. When using calcium and vitamin D preparations, pay attention to interactions with other drugs. For example, certain phosphate-lowering drugs may affect the absorption and effect of calcium preparations. Regularly monitor blood calcium levels and concurrently used active vitamin D, calcimimetics, etc., and adjust calcium intake.
[0130] S250, the degree of renal impairment, can be assessed by estimating the glomerular filtration rate (eGFR) using the CKD-EPI formula:
[0131] [eGFR=141\times min(Scr,3.5)^{1.2097}\times[0.949-0.675\times log(Scr)]^{-1.2097}]
[0132] Among them, min(Scr,3.5) means that if the blood creatinine value is greater than 3.5 mg / dL, 3.5 mg / dL is used; otherwise, the actual value is used.
[0133] S260. Collect other data, such as body mass index, low hemoglobin level, etc.
[0134] S300, Lifestyle and Nutrition Management
[0135] S310, Nutrition and Diet Management, monitor diet. It is necessary to limit phosphorus intake, increase calcium intake, and maintain a diversified diet.
[0136] S320. To change your lifestyle, you need to quit smoking and drinking and ensure adequate sleep.
[0137] S330: Exercise and training: monitor the patient's exercise volume and duration to determine whether the exercise meets the requirements.
[0138] S400, Complications prevention and care, common complications of hypocalcemia include:
[0139] Neuromuscular system symptoms: Hypocalcemia increases the excitability of the neuromuscular system. Common symptoms include muscle cramps, hand and foot cramps, numbness around the mouth, numbness and tingling in the limbs, etc. In severe cases, laryngeal spasm and bronchospasm may occur, and even cause epileptic seizures.
[0140] Cardiovascular system symptoms: Hypocalcemia may cause arrhythmia, rapid heartbeat, cyanosis, etc. These symptoms may be related to the increased neuromuscular excitability caused by hypocalcemia.
[0141] Respiratory symptoms: Hypocalcemia may lead to apnea, especially in cases of severe laryngospasm and bronchospasm.
[0142] Digestive system symptoms: Hypocalcemia may also cause digestive system discomfort, such as nausea, vomiting, etc.
[0143] Skin and nail changes: Long-term hypocalcemia can cause symptoms such as dry, inelastic skin and brittle nails.
[0144] Central nervous system symptoms: Severe hypocalcemia may affect the brain, causing symptoms such as confusion, memory loss, delirium, depression, and hallucinations.
[0145] S500, Continuous Monitoring and Evaluation
[0146] S510, regular review. According to the review cycle set by the medical staff, remind patients to have regular review. At the same time, upload the review information for comparison with the previous information to verify the prediction of the digital twin model, further predict the condition, provide corresponding nursing guidance, and provide patients with corresponding health education information.
[0147] S520, Assessment:
[0148] S521. Evaluation of predictions and extrapolations of digital twin models:
[0149] 1) Input patient monitoring information, such as medical examination information, patient feedback information, patient active feedback information, etc., and then compare it with the information predicted by the digital twin model to find inaccuracies in the predicted information and then make improvements. For example, according to the digital twin model, if the patient maintains a good attitude and complies with the doctor's orders and cooperates with the treatment, the blood calcium content should be normal; but after examination, it is found that the blood calcium content is still low, which means that the digital twin model prediction has deviated; the system prompts the medical staff to check, and the medical staff judges that the calcium and vitamin D supplements are insufficient, so the calcium and vitamin D supplements are increased, and the blood calcium concentration range is predicted for the next examination, waiting for the next examination. In this way, through multiple continuous adjustments, the digital twin model can be closer to the actual situation of the patient, which greatly increases the accuracy of the prediction.
[0150] 2) The digital twin model multiplies the value of each data type input by the proportion of the data's impact on the treatment outcome, and then takes the percentage; for the prediction results corresponding to multiple data, the percentages of each data are added together to obtain the comprehensive percentage; if the percentage is greater than or equal to 100%, it is predicted that the situation predicted by the data will definitely occur; if it is greater than or equal to 50%, it means that the situation predicted by the data has a high probability of occurring; if it is less than 50%, it means that the situation predicted by the data has a low probability of occurring;
[0151] After the follow-up review, the patient's actual situation is input, and then the system compares it. If the predicted situation with a percentage greater than or equal to 100% does not occur, it means that there is an error in the prediction. At this time, the system reminds the medical staff or maintenance personnel to make corrections, thereby continuously improving the prediction accuracy. If a low-probability situation occurs multiple times (at least 5 times or three times in a row), it means that the situation should be a high-probability event or a situation that should occur. At this time, the system also reminds the medical staff or maintenance personnel to make corrections.
[0152] S522. Evaluation of disease changes and nursing plans. Through the digital twin model, historical parameters, and revised parameters, the detected disease changes and nursing plans are output; cases where the treatment of the disease is not as effective as expected are pushed to medical staff, who will optimize or change the nursing plan and monitor the results of the implementation of the new plan. If the results are positive, the modified plan will be used as the preferred nursing plan and updated in the digital twin model to provide an optimization basis for subsequent care.
[0153] The above is only a preferred specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A low calcium care system for renal osteodystrophy in hemodialysis patients, characterized in that: include: The patient side is used to collect patient information, enable communication between the patient and the doctor, and obtain information from the server; The medical terminal is used for medical staff to upload patient data and obtain the data uploaded by the patient terminal and the corresponding analysis results from the server; The server is used to parse the data uploaded by the patient and doctor, and then store it in the ID directory identified by the ID data module, and process the corresponding data through the digital twin model and nursing module to output the results; Network, used to connect medical and nursing terminals, servers, and patients; ID data module, used to identify the ID to which the data belongs and perform data processing; The digital twin model collects a large amount of patient data, disease progression, and nursing plans to build a software model. Then, based on the specific patient data, adjustments made by the medical staff, and system identification, the software model is further optimized to become the digital twin model of the patient. The digital twin model is then used to predict disease progression and nursing plans. The nursing module is used to provide health education, social interaction, and virtual companion services to patients based on the data collected from patients by medical staff.
2. The low calcium care system according to claim 1, characterized in that: The software model comprises: Data collection, collecting patients’ physiological parameters and pathological data; Data processing: clean and assign values to the data. Ratio the measured value to the standard parameter value and retain two decimal places to obtain the assigned value. The measured parameter that is less than the minimum value of the standard parameter interval is ratioed to the minimum value of the standard parameter interval. The measured parameter that is greater than the maximum value of the standard parameter interval is ratioed to the maximum value of the standard parameter interval. The value between the minimum and maximum values of the standard parameter interval is 1. Weight some selected parameters by multiplying their assigned values by 2 to obtain the final assigned value. Data integration, which combines and fuses data from different data sources into a comprehensive and consistent patient data set; Feature extraction, selecting features relevant to medical decision-making from the integrated patient dataset; Modeling and training: using extracted features and medical expertise to build software models; Model verification and evaluation: treatment cases are selected and input into the model to verify the accuracy and predictive ability of the model, while the model performance is evaluated and patients are continuously optimized.
3. The low calcium care system according to claim 2, characterized in that: The digital twin model is obtained by relying on the software model and combining the selected parameter types; according to the selected parameter types, the model corresponding to this part of the parameter type is stripped from the software model and optimized to obtain the digital twin model.
4. The low calcium care system according to claim 2, characterized in that: The nursing module comprises: The semantic recognition module is used to recognize the semantics and perform corresponding operations according to the recognized semantics; The speech recognition module is used to recognize the input speech and convert it into text for semantic recognition or directly perform semantic recognition; An image recognition module is used to recognize input pictures and photos. The image recognition module is used to recognize input photos of medicines and food, so as to record the patient's medication and diet, or provide guidance on the patient's medication and diet; Movement recognition module, used to obtain the patient's movement status and movement amount; The status monitoring module is used for medication monitoring, diet monitoring, and health monitoring. Medication monitoring is based on the patient's own feedback and the photos of medication taken to monitor whether the patient is taking the medication according to the doctor's orders; diet monitoring monitors the patient's diet category by actively inputting food information through the patient's end and recognizing food information through images, and then calculates the impact of food on the patient based on the corresponding digital twin model to guide the patient to change his eating habits and eat healthily; health monitoring monitors the patient's health status through medical data and wearable data; The assistant module is used for health education, state adjustment, social adjustment, and habit inference. The patient can send a request to the server or the medical staff can actively push relevant popular knowledge to the patient through the server. The patient's mental state is judged through the information detected by the patient and the information actively input by the patient, and adjustment information is sent to the patient in a timely manner according to the corresponding state. The patient actively infers the patient's habits through the detected data, and then gives corresponding improvement or encouragement measures based on the patient's habits combined with the digital twin model. Virtual companions, combined with existing artificial intelligence technology, can initiate communication and companionship when the patient actively requests it or the system detects that the patient may need it, thereby further realizing psychological care.
5. A method for operating a low calcium system for renal osteodystrophy in hemodialysis patients, characterized in that: include: S100, Building a Digital Twin Model of Patients S110. According to the current research on the nursing of low calcium in hemodialysis patients with renal osteopathy, manually calibrate or select the corresponding data types through artificial intelligence technology, and then assign weights to each data type; S120. Separating the patient's digital twin model from the software model according to the selected data types; the selected data types in the digital twin model have been trained, and include the influence ratio of each data type on the treatment result and the weight of each data type; S130, predict and calculate based on the patient's data, and then compare with the actual results to make corrections and optimizations; S140, Health Belief Model S141. Educate patients. Output educational information to patients based on their conditions to help them understand their conditions, matters that require attention, lifestyle habits that need to be improved, and how to maintain a good attitude and increase compliance. S142, psychological support, assess the patient's mental state, and actively guide the patient to maintain a good attitude and confidence in the treatment effect; S143, monitoring and feedback, monitoring the patient's compliance, psychological stress and the impact of the health belief model on the development of the patient's condition through patient follow-up and patient self-test feedback, and continuously improving the deficiencies; S150, the digital twin model combines the input data to give treatment, nursing suggestions, and HBM suggestions; S200, treatment and monitoring, output treatment suggestions, nursing suggestions, and patient suggestions based on the digital twin model, and then medical staff specify treatment methods, nursing methods, HBM care and suggestions for patients based on the suggestions; use the digital twin model to infer the patient's disease progression and treatment effect based on the treatment method, nursing method, patient's pathological data, and patient's trial data, and regularly monitor the patient's actual parameters; S300, lifestyle and nutrition management, outputs recommendations on patients’ lifestyle and nutrition management based on the digital twin model to guide medical staff and patients to change their lifestyles and improve their nutrition according to the recommendations; S400, Complication Prevention and Care, predicts possible complications through the digital twin model, and takes preventive and nursing measures; the actual complications of the patient are input into the system to verify the accuracy of the digital twin model; S500, continuous monitoring and evaluation, continuously monitors the patient's data and compares it with the predictions of the digital twin model. If deviations are found, corrections are made to improve the prediction accuracy of the digital twin model.
6. The method for operating a low calcium system according to claim 5, characterized in that: In S110, the assignment process is as follows: if the data can be parameterized, the two decimal places after the ratio of the measured parameter to the standard parameter are used as the weight value, and then some parameters that need to be weighted are weighted to obtain the final assignment. The measured parameter that is less than the minimum value of the standard parameter interval is ratioed to the minimum value of the standard parameter interval, and the measured parameter that is greater than the maximum value of the standard parameter interval is ratioed to the maximum value of the standard parameter interval. The value between the minimum and maximum values of the standard parameter interval is 1. The HBM factor is a relatively subjective parameter, and the total score is assigned and each factor is evenly divided to assign the value.
7. The method for operating a low calcium system according to claim 5 or 6, characterized in that: In S120, the Cox proportional hazard model was used to verify whether the effect of each parameter on the result was significant through hypothesis testing. The parameters with significant effect on the result and the selected parameters were used to calculate their OR values through logistic regression, and the OR value was used as the impact ratio.
8. The method for operating a low calcium system according to claim 5, characterized in that: Also included in the S200: S210. Methods of monitoring dialysis treatment; S220, calcium and vitamin D supplementation, adjust the dose and frequency of oral calcium according to the patient's blood calcium level; regularly monitor blood calcium concentration, adjust the dose according to the results, and correct the digital twin model; S230, iPTH monitoring; S240, avoid drug interactions; S250. Assess the degree of renal function impairment by estimating glomerular filtration rate (eGFR).
9. The method for operating a low calcium system according to claim 5, characterized in that: The S300 also includes: S310, Nutrition and Diet Management, monitor diet, limit phosphorus intake, increase calcium intake, and maintain dietary diversity; S320. Change your lifestyle, quit smoking and drinking, and get enough sleep; S330: Exercise and training: monitor the patient's exercise volume and duration to determine whether the exercise meets the requirements.
10. The method for operating a low calcium system according to claim 5, characterized in that: The S500 also includes: S510, regular review: according to the review cycle set by the medical and nursing end, remind the patient to have a regular review, and upload the review information to compare with the previous information, verify the prediction of the digital twin model, further predict the condition, provide corresponding nursing guidance, and provide corresponding health education information to the patient; S520, Assessment: S521. Evaluation of predictions and extrapolations of digital twin models: 1) Input patient monitoring information and then compare it with the information predicted by the digital twin model to find inaccuracies in the predicted information and then make improvements; 2) The digital twin model multiplies the value of each data type input by the proportion of the data's impact on the treatment outcome, and then takes the percentage. For prediction results corresponding to multiple data, the percentages of each data are added together to obtain a comprehensive percentage. If the percentage is greater than or equal to 100%, it is predicted that the situation predicted by the data will definitely occur. If it is greater than or equal to 50%, it means that the situation predicted by the data is likely to occur. If it is less than 50%, it means that the situation predicted by the data is unlikely to occur. After the follow-up review, the patient's actual situation is input, and then the system compares it. If the predicted situation with a percentage greater than or equal to 100% does not occur, it means that the prediction is wrong. At this time, the system reminds the medical staff or maintenance personnel to make corrections; if a low-probability situation occurs multiple times, it means that the situation should be a high-probability event or a situation that should occur. At this time, the system also reminds the medical staff or maintenance personnel to make corrections; S522. Evaluation of disease changes and nursing plans. Through the digital twin model, historical parameters, and revised parameters, the detected disease changes and nursing plans are output; cases where the treatment of the disease is not as effective as expected are pushed to medical staff, who will optimize or change the nursing plan and monitor the results of the implementation of the new plan. If the results are positive, the modified plan will be used as the preferred nursing plan and updated in the digital twin model at the same time.
Citation Information
Cited By
Artificial intelligence-based nephropathy peritoneal dialysis full life cycle management device, system and method
CN119889587A
Digital twinning driven perioperative period nursing scheme dynamic optimization method
CN120878050A
Digital twin driven perioperative nursing plan dynamic optimization method
CN120878050B
