Deep neural network-based disease full-period management system

Abstract

The invention discloses a disease full-period management system based on a deep neural network, and the system comprises a collection time sequence planning module, a disease biomarker data collection module, and a compliance standard assessment module. The method comprises the following steps: acquiring previous disease diagnosis elements such as specific disease category proportion and attack frequency of each patient in each time window of a historical period from a target hospital patient platform, analyzing a disease period adaptive value according to the previous disease diagnosis elements, and further determining an acquisition time sequence plan when the patient is discharged from a hospital. And the compliance standard evaluation module is used for acquiring a compliance evaluation value of the discharge collection time sequence plan of each patient in the current period, judging whether the compliance reaches the standard or not, sending out an early warning prompt to the substandard patient, and recording the substandard patient as a to-be-analyzed patient. And the disease outcome state analysis module is used for analyzing the disease outcome states of the to-be-analyzed patients at each acquisition time point, carrying out early warning prompt according to the analysis result, and realizing accurate management and monitoring of the whole period of the diseases of the patients through multi-module cooperation.

Description

Technical Field

[0001] The present invention relates to the technical field of disease management, and in particular to a full-cycle disease management system based on a deep neural network. Background Art

[0002] As the medical industry moves toward digitalization, the amount of patient data accumulated by hospitals has exploded. How to use this data to optimize disease management has become an urgent problem to be solved. The traditional disease management model has many limitations. When judging the stage of a disease, doctors rely on experience and basic examinations, and it is difficult to accurately locate the stage of disease development, resulting in a lack of targeted treatment plans. In the rehabilitation stage, it is difficult to fully capture the microscopic changes in the patient's body and adjust the rehabilitation plan in time by relying solely on regular follow-up visits and routine physiological indicator tests. With the rise of deep learning technology, deep neural network algorithms have great potential for application in the medical field. It can deeply analyze patient data, accurately determine the stage of the disease, and formulate personalized data collection plans. Therefore, a disease full-cycle management system based on deep neural networks is needed.

[0003] Prior art, such as a disease data collection and analysis method and collection and analysis system disclosed in the invention application patent with announcement number: CN115919286B. On the one hand, by analyzing the internal nasal state evaluation index of the target patient corresponding to each collection time point, the accuracy and reliability of the target patient's corresponding internal nasal state evaluation results are guaranteed to the greatest extent, and the persuasiveness of the target patient's corresponding internal nasal state analysis results is improved, providing accurate and intuitive data for the subsequent target patient's corresponding rhinitis severity analysis. On the other hand, by analyzing the respiratory state evaluation index of the target patient corresponding to each collection time point, the singleness and one-sidedness of the patient's corresponding rhinitis severity monitoring and analysis in the current technology are effectively solved, and the reliability and accuracy of the patient's rhinitis severity analysis are increased.

[0004] In view of the above scheme, the inventors of the present application have found that the above technology has at least the following technical problems: 1. The disease staging of the prior art mainly relies on the clinical experience of doctors and traditional diagnostic methods. For example, for some complex diseases, doctors may judge the staging only based on the severity of symptoms and the time of onset. This subjective judgment is easily limited by the doctor's personal experience and knowledge level, which may lead to differences in the staging judgments of different doctors for the same patient. It may not be possible to fully integrate and analyze the previous disease diagnosis elements, which makes it impossible to fully consider the various characteristics of the disease during the disease staging process, thereby reducing the accuracy of the staging. Accurate staging is crucial for the selection of subsequent treatment plans. Inaccurate staging may lead to inaccurate treatment plans and affect the treatment effect. At the same time, there is a lack of support for deep neural network algorithms, and it is impossible to use a large amount of data to optimize the disease management model. This makes it difficult to continuously improve and improve the accuracy of disease staging, acquisition timing planning, and outcome evaluation, and it is impossible to give full play to the value of data to promote the advancement of medical technology.

[0005] Existing technologies may mainly rely on traditional clinical indicators when evaluating disease prognosis, such as a few markers in blood tests, imaging test results, etc. This evaluation method is relatively single and cannot comprehensively consider multi-dimensional information. Micro-level data is very important for early warning and accurate evaluation of disease prognosis. The lack of them will lead to inaccurate evaluation and difficulty in conducting continuous and dynamic disease prognosis evaluation of patients. Usually, examinations are performed at fixed time points to judge the prognosis, and data cannot be obtained in real time or frequently to track subtle changes in the disease prognosis process. This may lead to failure to detect problems in disease prognosis (such as slow or stagnant prognosis) in time, thereby missing the best time for intervention. At the same time, the patient's priority is not dynamically adjusted according to the patient's real-time disease prognosis status and data collection compliance. This may lead to unreasonable allocation of medical resources, for example, too many resources are allocated to patients with good prognosis, while patients with slow prognosis or poor data collection compliance do not receive timely attention and resource support.

[0006] 3. Existing technologies do not conduct quantitative assessments on data collection compliance, and it is difficult to accurately determine whether patients have collected data as required after discharge. As a result, it is not possible to promptly detect that patient data collection does not meet the standards. When problems occur during data collection, such as untimely collection or poor quality of collected samples, existing technologies are difficult to promptly detect and provide guidance and intervention to patients. This may result in missing or poor quality data, affecting the assessment of disease outcomes and subsequent treatment decisions. Summary of the invention

[0007] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to provide a disease full-cycle management system based on deep neural networks.

[0008] In order to solve the above technical problems, the present invention adopts the following technical solutions: The present invention provides a disease full-cycle management system based on a deep neural network, including: an acquisition timing planning module: used to obtain the previous disease diagnosis elements corresponding to each patient in the target hospital patient platform in each time window of the historical period, the previous disease diagnosis elements include the proportion of specific disease categories, the attack frequency, duration and staging numbers corresponding to each single disease, so as to analyze the disease stage adaptation value corresponding to each patient, and then analyze the acquisition timing planning of each patient when discharged from the hospital.

[0009] Compliance compliance assessment module: used to obtain the compliance assessment value corresponding to the discharge data collection timing plan of each patient in the current period, and then evaluate the compliance of the discharge data collection timing plan of each patient in the current period. If the compliance of a patient's discharge data collection timing plan does not meet the standard, an early warning will be issued, and each patient whose compliance with the discharge data collection timing plan does not meet the standard will be recorded as a high-risk patient.

[0010] Disease prognosis status analysis module: used to analyze the disease prognosis status of each high-risk patient at each collection time point, and provide early warning prompts based on the disease prognosis status of each high-risk patient at each collection time point.

[0011] The beneficial effects of the present invention are as follows: 1. In the embodiment of the present invention, a deep neural network algorithm is used to combine historical disease diagnosis parameters to more accurately divide the patient's disease stage. This helps doctors to formulate personalized treatment plans for patients that are more in line with their disease progression, rather than using traditional, relatively extensive disease staging methods. Based on the disease stage adaptation value matching corresponding discharge collection timing planning, the data collection plan for each patient is closely related to his or her disease state. More targeted data support is provided for subsequent treatment adjustments. Disease prognosis assessment is performed using microbiome and human symbiotic ecological data and intercellular communication molecular signal data, taking into account multi-dimensional information from microbial community structure to cell molecular level. These data can more sensitively reflect subtle changes in the patient's body, thereby more accurately judging whether the disease prognosis state is good, slow or stagnant, and providing a basis for doctors to adjust rehabilitation plans in a timely manner. 2. In the embodiment of the present invention, an early warning is issued in a timely manner according to the interval in which the disease prognosis assessment value is located, and a prompt can be given at an early stage when a problem with the disease prognosis occurs. Avoid further deterioration of the condition and improve the treatment effect. Evaluating and warning the compliance of the patient's discharge collection timing planning can ensure the quality and integrity of data collection. Timely discovery of patient problems during the data collection process also helps to timely discover possible non-cooperation or forgetfulness of patients, timely intervention, and ensure the smooth progress of data collection. In the background of the medical management system, the patient priority is dynamically adjusted according to the patient's disease prognosis status and data collection compliance. For patients with good prognosis, the priority is appropriately lowered so that medical resources can be allocated more to patients with poor prognosis or poor data collection compliance, thereby improving the utilization efficiency of medical resources. When patients encounter emergency situations such as stagnation of disease prognosis, the system can quickly raise the patient's condition to the highest priority, timely deploy medical resources, better respond to the patient's emergency medical needs, and improve the overall quality of medical services. 3. In the embodiment of the present invention, the system records the patient's good prognosis and other data, which can be used as a reference for subsequent rehabilitation cases of similar diseases. By accumulating a large amount of patient data and treatment experience, it helps doctors better understand the prognosis of the disease, provide a more scientific basis for the treatment of future patients, and promote the accumulation and sharing of medical knowledge. As the system collects more and more data, including the patient's disease parameters, prognosis, and data collection compliance, these data can be used to further optimize the deep neural network model. This enables the accuracy and reliability of the model to be continuously improved in various aspects such as disease stage analysis, collection timing planning, disease outcome assessment and compliance assessment, forming a virtuous data-model optimization cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0013] Figure 1 This is a schematic diagram of the connection of the system modules of the present invention. DETAILED DESCRIPTION

[0014] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0015] Embodiments of the present invention Figure 1 As shown, a disease full-cycle management system based on deep neural network includes: an acquisition timing planning module, a compliance achievement assessment module, a disease outcome status analysis module and a database.

[0016] The compliance achievement assessment module is connected to the acquisition timing planning module and the disease outcome status analysis module respectively, and the database is connected to the acquisition timing planning module and the disease outcome status analysis module.

[0017] Collection timing planning module: used to obtain the previous disease diagnosis elements corresponding to each patient in the target hospital patient platform in each time window of the historical period. The previous disease diagnosis elements include the proportion of specific disease categories, the attack frequency, duration and stage numbers corresponding to each single disease, so as to analyze the disease stage adaptation value corresponding to each patient, and then analyze the collection timing planning of each patient when discharged from the hospital.

[0018] In a specific embodiment, the disease stage adaptation value corresponding to each patient is analyzed, and the specific analysis process is as follows: the proportion of specific disease categories corresponding to each patient in each time window of the historical period, the attack frequency, duration and stage number corresponding to each single disease are recorded as , , and ,in, Indicates the number corresponding to each patient, , q is the sum of all patients, q is any integer greater than 2, Indicates the number corresponding to each time window, , r is the sum of all time windows, r is also any integer greater than 2, Indicates the number corresponding to each single disease, , w is the sum of all single diseases, w is also any integer greater than 2, substitute into the calculation formula The disease stage adaptation value corresponding to each patient is obtained ,in, , , , They are the standard specific disease category proportion corresponding to the set patients, the standard attack frequency corresponding to a single disease, the standard duration, and the standard staging number. , , , They are the weight factors corresponding to the proportion of specific disease categories of patients, the weight factors corresponding to the frequency of attacks of a single disease, the weight factors corresponding to the duration, and the weight factors corresponding to the stage numbers. Represents a natural constant.

[0019] It should be noted that , , , Both are greater than 0 and less than 1.

[0020] It should also be noted that through the summary of a large amount of research data and experimental data, professional institutions and research institutions set the standard specific disease category proportion corresponding to patients, the standard attack frequency corresponding to a single disease, the standard duration, and the standard staging number. At the same time, based on the professional knowledge and research basis of experts in the field, they discuss and confirm with industry organizations or professional institutions. Experts set the weight factors corresponding to the proportion of patients' specific disease categories, the weight factors corresponding to the attack frequency of a single disease, the weight factors corresponding to the duration, and the weight factors corresponding to the staging numbers based on their own experience and knowledge.

[0021] Once again, it is necessary to explain that the disease diagnosis records in the hospital information system are classified and counted. The system records the name of the disease diagnosed each time the patient visits the doctor. By writing a data extraction program, the number of times each disease appears in all the patient's diagnosis records within a specific historical period can be counted. Then, the proportion of the number of times each disease appears in the total number of diagnoses is calculated to obtain the proportion of specific disease categories. In the hospital information system, the date of the visit can be used as a reference for the onset time. If a patient visits the doctor for the same disease multiple times, the reciprocal of the time interval between two adjacent visits can be calculated to obtain a preliminary estimate of the frequency of onset. The time span from the first time the patient was diagnosed with a certain disease to the last record showing that the disease still exists or is cured is the duration of the disease.

[0022] In a specific embodiment, the acquisition timing plan of each patient at the time of discharge is analyzed, and the specific analysis process is as follows: the disease stage adaptation value corresponding to each patient is compared with the disease stage adaptation value corresponding to each acquisition timing plan in the database. If the disease stage adaptation value corresponding to a patient is the same as the disease stage adaptation value corresponding to a certain acquisition timing plan in the database, then the acquisition timing plan in the database is recorded as the acquisition timing plan corresponding to the patient's discharge.

[0023] The collection timing plan includes the collection time points corresponding to each patient after discharge, and lists the collection time points corresponding to each patient after discharge into a table, thereby forming a collection form corresponding to each patient, and then sending the collection form corresponding to each patient to each patient's communication device.

[0024] Compliance compliance assessment module: used to obtain the compliance assessment value corresponding to the discharge data collection timing plan of each patient in the current period, and then evaluate the compliance of the discharge data collection timing plan of each patient in the current period. If the compliance of a patient's discharge data collection timing plan does not meet the standard, an early warning will be issued, and each patient whose compliance with the discharge data collection timing plan does not meet the standard will be recorded as a high-risk patient.

[0025] In a specific embodiment, the compliance evaluation value corresponding to the discharge collection timing plan of each patient in the current period is obtained, and the specific acquisition process is as follows: the sample collection quantity, collection coverage, and sample collection time deviation corresponding to the discharge collection timing plan of each patient in the current period are obtained, and the sample collection quantity, collection coverage, and sample collection time deviation corresponding to the discharge collection timing plan of each patient in the current period are recorded as , and ,in, Indicates the number corresponding to each patient, , q is any integer greater than 2, substitute it into the calculation formula: The compliance evaluation value corresponding to the discharge collection timing plan of each patient in the current period is obtained ,in, , and They are the number of standard sample collections, the standard collection coverage, and the standard sample collection time deviation corresponding to the set patient discharge collection timing plan. , , They are the weight factors corresponding to the number of samples collected in the set patient discharge collection timing plan, the weight factors corresponding to the collection coverage, and the weight factors corresponding to the sample collection time deviation.

[0026] It should be noted that , , Both are greater than 0 and less than 1.

[0027] It should also be noted that through the summary of a large amount of research data and experimental data. According to professional institutions and research institutions, the standard sample collection quantity, standard collection coverage, and standard sample collection time deviation corresponding to the patient discharge collection timing plan are set. At the same time, based on the professional knowledge and research basis of experts in the field, they are discussed and confirmed with industry organizations or professional institutions. Experts set the weight factors corresponding to the sample collection quantity, the collection coverage, and the sample collection time deviation of the patient discharge collection timing plan based on their own experience and knowledge.

[0028] Once again, it is important to note that electronic health record (EHR) systems or specialized data collection software are used, which typically record information about each patient sample collection. When a patient completes a collection, they enter the type of sample (such as blood, feces, tissue, etc.) and quantity collected into the system, record the number of samples collected and the collection coverage, and compare the actual collection time with the standard collection time (or time window) to obtain the time deviation.

[0029] In a specific embodiment, the compliance corresponding to the discharge collection timing plan of each patient in the current period is evaluated, and the specific evaluation process is as follows: the compliance evaluation value corresponding to the discharge collection timing plan of each patient in the current period is compared with the compliance evaluation value corresponding to the set standard patient discharge collection timing plan. If the compliance evaluation value corresponding to the discharge collection timing plan of a patient in the current period is greater than or equal to the set compliance evaluation value interval of the standard patient discharge collection timing plan, then the compliance corresponding to the discharge collection timing plan of the patient in the current period is evaluated to be up to standard; if the compliance evaluation value corresponding to the discharge collection timing plan of a patient in the current period is less than the set compliance evaluation value interval of the standard patient discharge collection timing plan, then the compliance corresponding to the discharge collection timing plan of the patient in the current period is evaluated to be not up to standard. In this way, the compliance corresponding to the discharge collection timing plan of each patient in the current period is evaluated.

[0030] In a specific embodiment, when the compliance corresponding to the discharge collection sequence plan of a certain patient does not meet the standard, an early warning prompt will be issued. The specific early warning process is as follows: the system will quickly send an early warning message to the patient and his / her pre-designated family members through the SMS platform. The content of the SMS informs the patient that there is a problem with the execution of the current collection plan. At the same time, it is attached with a consultation telephone number specially established by the hospital. Within the hospital, the medical information management system will simultaneously generate a striking early warning task reminder for the medical team responsible for the patient. The reminder will be displayed in a prominent form on the homepage of the medical staff's work interface to ensure that the medical staff can notice it at the first time. When the medical staff clicks on the reminder, they can immediately view the detailed early warning report. The report lists in detail the specific data of the non-compliance and the degree of deviation from the standard. At the same time, the patient's past data collection history records will be displayed in series. After receiving the early warning, the medical staff must take the initiative to contact the patient or family members strictly within the specified time. During the communication process, the medical staff will patiently and carefully ask about various problems encountered during the data collection process.

[0031] Disease prognosis status analysis module: used to analyze the disease prognosis status of each high-risk patient at each collection time point, and provide early warning prompts based on the disease prognosis status of each high-risk patient at each collection time point.

[0032] In a specific embodiment, the disease outcome status analysis module also includes a data acquisition unit and a disease outcome assessment value analysis unit.

[0033] The data acquisition unit is used to collect data according to the corresponding collection timing plan when each high-risk patient is discharged from the hospital, so as to obtain the microbiome and human symbiotic ecological data and intercellular communication molecular signal data corresponding to each high-risk patient at each collection time point. The microbiome and human symbiotic ecological data include the number of bacterial species corresponding to each part, the proportion of dominant bacterial species and the microbial community diversity index, and the intercellular communication molecular signal data include the EVs concentration corresponding to each part, the EVs-carrying molecular content and the neurotransmitter concentration, and then analyze and obtain the microbiome and human symbiotic ecological fitness value and intercellular communication molecular signal fitness value corresponding to each high-risk patient at each collection time point.

[0034] It should be noted that patients can purchase and operate the convenient home microbial testing kits themselves according to the instructions. These kits usually include sampling tools, such as oral swabs, skin swabs, stool collection tubes, etc. and testing reagents. The microbial testing kits will be equipped with a smartphone application, which will scan the results of the reaction between the testing reagents and samples, such as color changes and fluorescence intensity, through the mobile phone camera. The APP uses a built-in algorithm to convert these signals into preliminary microbial information. The APP counts the number of bacterial species based on the built-in microbial database, and determines the dominant bacterial species and their proportion and the microbial community diversity index by comparing the signal strength corresponding to different bacterial species. .

[0035] It should also be noted that patients use a microfluidic home test kit for testing, collecting a small blood sample. The blood sample is collected into a matching micro-blood collection tube, which is pre-coated with an anticoagulant to prevent blood clotting. The blood sample in the blood collection tube is dripped onto a specially designed home microfluidic chip. The chip has tiny channels and reaction chambers inside, and the channels contain nanoparticles labeled with antibodies or aptamers that can specifically bind to EVs. When blood flows through the channels, EVs bind to these labeled nanoparticles. Simple optical detection devices such as light-emitting diodes and photodetectors are integrated on the chip. When the EVs-nanoparticle complex passes through the detection area, it causes changes in light scattering or fluorescence signals. Through the mobile phone application, the built-in algorithm converts the light signal changes into EVs concentration values ​​and displays them on the mobile phone screen in real time. At the same time, the channels of the internal device of the microfluidic home test kit contain special reagents. When EVs flow through the channels under capillary action, the reagents can cause EVs to rupture and release the molecules they carry. In specific areas of the channel, probes that can specifically bind to target molecules are pre-fixed. When the molecule binds to the probe, a color change or fluorescent signal will be generated. The patient can use the mobile phone camera to capture the signal area and analyze the image through the accompanying mobile phone application. The application obtains the content of molecules carried by EVs based on the color intensity or fluorescence intensity and refers to a pre-established standard curve (made in the laboratory with a known concentration of molecular samples). A special skin patch sensor is used, which contains materials that can selectively adsorb neurotransmitters and electrochemical detection electrodes. The patient sticks the patch on the surface of the skin, such as the inside of the wrist or the forehead, and the neurotransmitters on the skin surface enter the patch through the micro-osmosis of the skin. The electrodes on the patch are connected to a small portable electrochemical detector, which can monitor the changes in electrochemical signals on the electrode surface in real time. When neurotransmitters bind to the adsorbent material and undergo redox reactions on the electrode surface, current changes are generated. The detector converts the current changes into neurotransmitter concentrations based on a pre-calibrated standard curve (established in the laboratory with a known concentration of neurotransmitter solution) and displays them on the screen of the detector.

[0036] Once again, it is important to note that there are already some functional products on the market. For example, there is a home microbial testing kit called "MicrobeScope", which is equipped with a microscope attachment that is compatible with mobile phones. Users can use the mobile phone camera to take images of microbial samples, and then use the accompanying mobile phone application for image analysis. The application has built-in machine learning algorithms that can identify common microbial species and provide some basic microbial characteristic information. There is also a product called "VesicleDetect", which is mainly aimed at EVs detection. It uses a special immune capture technology combined with mobile phone camera scanning detection. The user adds the sample to the detection device. After specific treatment, the device will cause the specific molecules carried by EVs to bind to antibodies and generate visual signals. After scanning with the mobile phone camera, the application analyzes parameters such as signal intensity, and then estimates the relative content of molecules carried by EVs, providing users with test results about EVs.

[0037] The disease outcome assessment value analysis unit is used to analyze and obtain the disease outcome assessment value corresponding to each high-risk patient at each collection time point based on the microbiome-human symbiotic ecological fitness value and the intercellular communication molecular signal fitness value corresponding to each high-risk patient at each collection time point.

[0038] In a specific embodiment, the analysis obtains the microbiome-human symbiotic ecological fitness value and the intercellular communication molecular signal fitness value corresponding to each high-risk patient at each collection time point. The specific analysis process is as follows: A1. The number of bacterial species, the proportion of dominant bacterial species and the microbial community diversity index corresponding to each part of each high-risk patient at each collection time point are input into the microbiome-human symbiotic ecological fitness value analysis model, and the microbiome-human symbiotic ecological fitness value corresponding to each high-risk patient at each collection time point is output.

[0039] It should be noted that the analysis process of the adaptation value of the microbiome and human symbiotic ecology corresponding to each high-risk patient at each collection time point is as follows: the number of bacterial species, the proportion of dominant bacterial species and the microbial community diversity index corresponding to each part of each high-risk patient at each collection time point are normalized, and the number of bacterial species, the proportion of dominant bacterial species and the microbial community diversity index corresponding to each part of each high-risk patient at each collection time point after processing are recorded as , and , substitute into the analysis formula , and obtain the symbiotic ecological adaptation value between the microbiome and the human body corresponding to each high-risk patient at each collection time point .

[0040] A2. Input the EVs concentration, EVs-carrying molecule content and neurotransmitter concentration corresponding to each part of each high-risk patient at each collection time point into the intercellular communication molecule signal adaptation value analysis model, and output the intercellular communication molecule signal adaptation value corresponding to each high-risk patient at each collection time point.

[0041] It should be noted that the analysis process of the cell-to-cell communication molecular signal adaptation value corresponding to each high-risk patient at each collection time point is as follows: the EVs concentration, EVs-carrying molecular content and neurotransmitter concentration corresponding to each part of each high-risk patient at each collection time point are normalized, and the EVs concentration, EVs-carrying molecular content and neurotransmitter concentration corresponding to each part of each high-risk patient at each collection time point after treatment are recorded as , and , substitute into the analysis formula , and obtain the cell-to-cell communication molecular signal adaptation value corresponding to each high-risk patient at each collection time point .

[0042] In a specific embodiment, the analysis obtains the disease outcome assessment value corresponding to each high-risk patient at each collection time point, and the specific analysis process is as follows: the microbiome and human symbiotic ecological fitness value and the intercellular communication molecular signal fitness value corresponding to each high-risk patient at each collection time point are recorded as and ,in, Indicates the number corresponding to each high-risk patient, , p is any integer greater than 2, Indicates the number corresponding to each collection time point, , v is any integer greater than 2, substitute it into the calculation formula: The disease outcome assessment values ​​corresponding to each high-risk patient at each collection time point are obtained. ,in, and They are the standard microbiome and human symbiotic ecological adaptation values ​​and standard intercellular communication molecular signal adaptation values ​​corresponding to the patient's discharge. , They are the weight factors corresponding to the adaptation value of the patient's discharged microbiome and the human symbiotic ecology, and the weight factors corresponding to the adaptation value of the intercellular communication molecular signal. , They are the regulatory factors corresponding to the set proportion of the patient's discharged microbiome and the human symbiotic ecological fitness value, and the regulatory factors corresponding to the proportion of the intercellular communication molecular signal fitness value. , They are respectively the regulatory factors corresponding to the set adaptation values ​​of the patient's discharged microbiome and the human symbiotic ecology, and the regulatory factors corresponding to the adaptation values ​​of the intercellular communication molecular signals.

[0043] It should be noted that , Both are greater than 0 and less than 1.

[0044] It should also be noted that through the summary of a large amount of research data and experimental data. According to professional institutions and research institutions, the standard microbiome and human symbiotic ecological fitness values ​​and standard intercellular communication molecular signal fitness values ​​corresponding to the patient's discharge are set. At the same time, based on the professional knowledge and research basis of experts in the field, they are discussed and confirmed with industry organizations or professional institutions. Experts set the weight factors corresponding to the fitness values ​​of the patient's discharged microbiome and the human symbiotic ecology, the weight factors corresponding to the fitness values ​​of the intercellular communication molecular signals, and the adjustment factors corresponding to the proportions of the patient's discharged microbiome and the human symbiotic ecology fitness values, and the adjustment factors corresponding to the proportions of the intercellular communication molecular signal fitness values, as well as the adjustment factors corresponding to the patient's discharged microbiome and the human symbiotic ecology fitness values, and the adjustment factors corresponding to the fitness values ​​of the intercellular communication molecular signals.

[0045] Once again, it should be noted that the regulatory factors corresponding to the self-ratio of the fitness value of the microbiome and human symbiotic ecology at discharge and the self-ratio of the fitness value of the molecular signal of intercellular communication are mainly used to emphasize or weaken the impact of the changes in the self-ratio of the fitness value of the microbiome and human symbiotic ecology and the fitness value of the molecular signal of intercellular communication on the disease outcome assessment value. The self-ratio refers to the ratio between the actual fitness value and the standard fitness value. or When it is greater than 1, it means that the change in the proportion of the symbiotic ecological fitness value between the microbiome and the human body or the fitness value of the molecular signal of intercellular communication itself has an amplified impact on the disease outcome assessment value.

[0046] It should be noted once again that the regulatory factors corresponding to the symbiotic ecological fitness value of the patient's discharge microbiome and the human body and the regulatory factors corresponding to the intercellular communication molecular signal fitness value take into account the comprehensive impact of the deviation of the microbiome and human symbiotic ecological fitness value and the intercellular communication molecular signal fitness value from the standard value on the disease outcome assessment value. This degree of deviation does not only depend on the proportional relationship. For example, the impact of the absolute value of the difference on the assessment value will be considered. or When it is greater than 0, the greater the deviation between the adaptation value and the standard value, the greater the impact on the disease outcome assessment value.

[0047] In a specific embodiment, the early warning is performed according to the disease prognosis status corresponding to each high-risk patient at each collection time point, and the specific early warning process is as follows: B1, the disease prognosis evaluation value corresponding to each high-risk patient at each collection time point is compared with the disease prognosis evaluation value interval corresponding to each disease prognosis status in the database. If the disease prognosis evaluation value corresponding to a certain high-risk patient at a certain collection time point is within the disease prognosis evaluation value interval corresponding to a certain disease prognosis status in the database, then the disease prognosis status in the database is recorded as the disease prognosis status corresponding to the high-risk patient at the collection time point. The disease prognosis status includes good prognosis, slow prognosis and stagnant prognosis; B2. If the disease outcome of a high-risk patient at a certain collection time point is good, a positive early warning message is generated to inform the high-risk patient and his / her responsible doctor that the current disease outcome is progressing smoothly and that the existing rehabilitation plan can continue to be implemented. At the same time, regular review is reminded to continuously monitor the outcome. In addition, in the background of the medical management system, the priority of the patient is lowered, and the good outcome is recorded as a reference for subsequent rehabilitation cases of similar diseases; It should be noted that the content of the early warning message is: "Dear [patient's name], according to the current collected data evaluation, your disease prognosis at [specific collection time point] is good, which shows that your current rehabilitation plan is very effective. Please continue to maintain it. Remember to return to the hospital regularly for checkups as ordered by your doctor. I wish you a speedy and complete recovery."

[0048] B3. If the disease progression status of a high-risk patient at a certain collection time point is slow progression, a corresponding early warning prompt will be generated to remind the high-risk patient and his / her responsible doctor that the current disease progression is slightly slower than expected, and it is recommended to make appropriate adjustments to the rehabilitation plan. At the medical system level, the patient will be marked as a key focus, and medical staff will be arranged to make regular follow-up visits to understand the situation. At the same time, a multidisciplinary medical team will be convened to consult on the patient's rehabilitation plan; It should be noted that the warning message reads: "Dear [patient's name], through the evaluation of the latest collected data, it was found that your disease progression at [specific collection time point] is slow, but please don't worry too much. This may be caused by multiple factors. We recommend that you communicate further with your doctor and consider whether you need to make appropriate optimizations to your current rehabilitation plan to better promote physical progression."

[0049] B4. If the disease progression status of a high-risk patient at a certain collection time point is progression stagnation, an emergency warning will be issued to remind the high-risk patient and the doctor that the current disease progression has stagnated and emergency intervention measures must be taken as soon as possible. At the same time, in the hospital's internal medical management process, the patient's condition will be given the highest priority, relevant medical resources will be quickly deployed, expedited examinations will be arranged, authoritative experts will be invited to participate in consultations, and the relevant data of the progression stagnation and subsequent measures will be recorded in detail.

[0050] It should be noted that the warning message reads: "Dear [patient's name], we are very sorry to inform you that according to the latest disease progression assessment results, your disease progression at [specific collection time point] has stagnated. This means that the physical progression is not proceeding as expected and there may be more serious obstacles. Please be sure to contact your attending physician as soon as possible. We will immediately arrange a comprehensive review and expert consultation for you, re-evaluate your condition and develop new, more targeted treatment and rehabilitation plans. We hope to help you break the stagnation of progression and return to health as soon as possible."

[0051] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they shall all fall within the protection scope of the present invention.

Claims

1. A disease full-cycle management system based on deep neural network, characterized in that: include: Acquisition timing planning module: used to obtain the previous disease diagnosis elements corresponding to each patient in the target hospital patient platform in each time window of the historical period. The previous disease diagnosis elements include the proportion of specific disease categories, the frequency of attacks, duration and stage numbers corresponding to each single disease, so as to analyze the disease stage adaptation value corresponding to each patient, and then analyze the acquisition timing planning when each patient is discharged from the hospital; Compliance compliance assessment module: used to obtain the compliance assessment value corresponding to the discharge collection timing plan of each patient in the current period, and then evaluate the compliance of the discharge collection timing plan of each patient in the current period. If the compliance of a patient's discharge collection timing plan does not meet the standard, an early warning prompt will be issued, and each patient whose discharge collection timing plan compliance does not meet the standard will be recorded as a high-risk patient; Disease prognosis status analysis module: used to analyze the disease prognosis status of each high-risk patient at each collection time point, and provide early warning prompts based on the disease prognosis status of each high-risk patient at each collection time point.

2. A disease full cycle management system based on deep neural network as claimed in claim 1, characterized in that: The disease stage adaptation value corresponding to each patient is analyzed, and the specific analysis process is as follows: The proportion of specific disease categories corresponding to each patient in each time window of the historical period, the frequency of onset, duration and stage number of each single disease are recorded as , , and ,in, Indicates the number corresponding to each patient, , q is the sum of all patients, q is any integer greater than 2, Indicates the number corresponding to each time window, , r is the sum of all time windows, r is also any integer greater than 2, Indicates the number corresponding to each single disease, , w is the sum of all single diseases, w is also any integer greater than 2, substitute into the calculation formula The disease stage adaptation value corresponding to each patient is obtained ,in, , , , They are the standard specific disease category proportion corresponding to the set patients, the standard attack frequency corresponding to a single disease, the standard duration, and the standard staging number. , , , They are the weight factors corresponding to the proportion of specific disease categories of patients, the weight factors corresponding to the frequency of attacks of a single disease, the weight factors corresponding to the duration, and the weight factors corresponding to the stage numbers. Represents a natural constant.

3. A disease full-cycle management system based on deep neural network as claimed in claim 2, characterized in that: The collection timing plan of each patient when discharged from the hospital is analyzed, and the specific analysis process is as follows: Compare the disease stage adaptation value corresponding to each patient with the disease stage adaptation value corresponding to each acquisition timing plan in the database. If the disease stage adaptation value corresponding to a patient is the same as the disease stage adaptation value corresponding to a certain acquisition timing plan in the database, then record the acquisition timing plan in the database as the acquisition timing plan corresponding to the patient's discharge; The collection timing plan includes the collection time points corresponding to each patient after discharge, and lists the collection time points corresponding to each patient after discharge into a table, thereby forming a collection form corresponding to each patient, and then sending the collection form corresponding to each patient to each patient's communication device.

4. A disease full cycle management system based on deep neural network as claimed in claim 3, characterized in that: The specific acquisition process of obtaining the compliance evaluation value corresponding to the discharge collection timing plan of each patient in the current period is as follows: Obtain the sample collection quantity, collection coverage, and sample collection time deviation corresponding to the discharge collection timing plan of each patient in the current period, and record the sample collection quantity, collection coverage, and sample collection time deviation corresponding to the discharge collection timing plan of each patient in the current period as , and ,in, Indicates the number corresponding to each patient, , q is any integer greater than 2, substitute it into the calculation formula: The compliance evaluation value corresponding to the discharge collection timing plan of each patient in the current period is obtained ,in, , and They are the number of standard sample collections, the standard collection coverage, and the standard sample collection time deviation corresponding to the set patient discharge collection timing plan. , , They are the weight factors corresponding to the number of samples collected in the set patient discharge collection timing plan, the weight factors corresponding to the collection coverage, and the weight factors corresponding to the sample collection time deviation.

5. A disease full-cycle management system based on deep neural network as claimed in claim 4, characterized in that: The evaluation process of the compliance of the discharge collection timing plan of each patient in the current period is as follows: The compliance assessment value corresponding to the discharge collection timing plan of each patient in the current period is compared with the compliance assessment value corresponding to the set standard patient discharge collection timing plan. If the compliance assessment value corresponding to the discharge collection timing plan of a patient in the current period is greater than or equal to the compliance assessment value interval corresponding to the standard patient discharge collection timing plan, then it is assessed that the compliance of the discharge collection timing plan of the patient in the current period is up to standard; if the compliance assessment value corresponding to the discharge collection timing plan of a patient in the current period is less than the compliance assessment value interval corresponding to the standard patient discharge collection timing plan, then it is assessed that the compliance of the discharge collection timing plan of the patient in the current period is not up to standard. In this way, the compliance of the discharge collection timing plan of each patient in the current period is evaluated.

6. A disease full-cycle management system based on deep neural network as claimed in claim 5, characterized in that: When the compliance of the patient's discharge collection timing plan does not meet the standard, an early warning prompt is issued. The specific early warning process is as follows: The system will quickly send warning information to the patient and his / her pre-designated family members through the SMS platform. The content of the SMS will inform the patient that there is a problem with the execution of the current collection plan. At the same time, it will attach a consultation telephone number specially established by the hospital. Within the hospital, the medical information management system will simultaneously generate a striking warning task reminder for the medical team responsible for the patient. The reminder will be displayed in a prominent form on the homepage of the medical staff's work interface to ensure that the medical staff can notice it at the first time. When the medical staff clicks on the reminder, they can immediately view the detailed warning report. The report lists in detail the specific data of the non-compliance and the degree of deviation from the standard. At the same time, the patient's past data collection history records will be displayed in series. After receiving the warning, the medical staff must take the initiative to contact the patient or family members strictly within the specified time. During the communication process, the medical staff will patiently and carefully ask about various problems encountered in the data collection process.

7. A disease full cycle management system based on deep neural network as claimed in claim 6, characterized in that: The disease outcome status analysis module also includes a data acquisition unit and a disease outcome assessment value analysis unit; The data acquisition unit is used to collect data according to the corresponding collection timing plan when each high-risk patient is discharged from the hospital, so as to obtain the microbiome and human symbiotic ecological data and intercellular communication molecular signal data corresponding to each high-risk patient at each collection time point. The microbiome and human symbiotic ecological data include the number of bacterial species corresponding to each part, the proportion of dominant bacterial species and the microbial community diversity index, and the intercellular communication molecular signal data include the EVs concentration, EVs carrying molecular content and neurotransmitter concentration corresponding to each part, and then analyze and obtain the microbiome and human symbiotic ecological fitness value and intercellular communication molecular signal fitness value corresponding to each high-risk patient at each collection time point; The disease outcome assessment value analysis unit is used to analyze and obtain the disease outcome assessment value corresponding to each high-risk patient at each collection time point based on the microbiome-human symbiotic ecological fitness value and the intercellular communication molecular signal fitness value corresponding to each high-risk patient at each collection time point.

8. A disease full cycle management system based on deep neural network as claimed in claim 7, characterized in that: The analysis obtained the microbiome-human symbiotic ecological fitness value and intercellular communication molecular signal fitness value corresponding to each high-risk patient at each collection time point. The specific analysis process is as follows: A1. Input the number of bacterial species, the proportion of dominant bacterial species and the diversity index of the microbial community corresponding to each part of each high-risk patient at each collection time point into the microbiome and human symbiotic ecological fitness value analysis model, and output the microbiome and human symbiotic ecological fitness value corresponding to each high-risk patient at each collection time point; A2. Input the EVs concentration, EVs-carrying molecule content and neurotransmitter concentration corresponding to each part of each high-risk patient at each collection time point into the intercellular communication molecule signal adaptation value analysis model, and output the intercellular communication molecule signal adaptation value corresponding to each high-risk patient at each collection time point.

9. A disease full cycle management system based on deep neural network as claimed in claim 8, characterized in that: The analysis obtains the disease outcome assessment value corresponding to each high-risk patient at each collection time point. The specific analysis process is as follows: The adaptation value of the microbiome and human symbiotic ecology and the adaptation value of the intercellular communication molecular signal corresponding to each high-risk patient at each collection time point are recorded as and ,in, Indicates the number corresponding to each high-risk patient, , p is any integer greater than 2, Indicates the number corresponding to each collection time point, , v is any integer greater than 2, substitute it into the calculation formula: The disease outcome assessment values ​​corresponding to each high-risk patient at each collection time point are obtained. ,in, and They are the standard microbiome and human symbiotic ecological adaptation values ​​and standard intercellular communication molecular signal adaptation values ​​corresponding to the patient's discharge. , They are the weight factors corresponding to the adaptation value of the patient's discharged microbiome and the human symbiotic ecology, and the weight factors corresponding to the adaptation value of the intercellular communication molecular signal. , They are the regulatory factors corresponding to the set proportion of the patient's discharged microbiome and the human symbiotic ecological fitness value, and the regulatory factors corresponding to the proportion of the intercellular communication molecular signal fitness value. , They are respectively the regulatory factors corresponding to the set adaptation values ​​of the patient's discharged microbiome and the human symbiotic ecology, and the regulatory factors corresponding to the adaptation values ​​of the intercellular communication molecular signals.

10. A disease full cycle management system based on deep neural network as claimed in claim 9, characterized in that: According to the disease prognosis status of each high-risk patient at each collection time point, an early warning prompt is provided. The specific early warning process is as follows: B1. Compare the disease outcome assessment value corresponding to each high-risk patient at each collection time point with the disease outcome assessment value interval corresponding to each disease outcome status in the database. If the disease outcome assessment value corresponding to a high-risk patient at a certain collection time point is within the disease outcome assessment value interval corresponding to a disease outcome status in the database, then the disease outcome status in the database is recorded as the disease outcome status corresponding to the high-risk patient at the collection time point. The disease outcome status includes good outcome, slow outcome and stagnant outcome; B2. If the disease outcome of a high-risk patient at a certain collection time point is good, a positive early warning message is generated to inform the high-risk patient and his / her responsible doctor that the current disease outcome is progressing smoothly and that the existing rehabilitation plan can continue to be implemented. At the same time, regular review is reminded to continuously monitor the outcome. In addition, in the background of the medical management system, the priority of the patient is lowered, and the good outcome is recorded as a reference for subsequent rehabilitation cases of similar diseases; B3. If the disease progression status of a high-risk patient at a certain collection time point is slow progression, a corresponding early warning prompt will be generated to remind the high-risk patient and his / her responsible doctor that the current disease progression is slightly slower than expected, and it is recommended to make appropriate adjustments to the rehabilitation plan. At the medical system level, the patient will be marked as a key focus, and medical staff will be arranged to make regular follow-up visits to understand the situation. At the same time, a multidisciplinary medical team will be convened to consult on the patient's rehabilitation plan; B4. If the disease progression status of a high-risk patient at a certain collection time point is progression stagnation, an emergency warning will be issued to remind the high-risk patient and the doctor that the current disease progression has stagnated and emergency intervention measures must be taken as soon as possible. At the same time, in the hospital's internal medical management process, the patient's condition will be given the highest priority, relevant medical resources will be quickly deployed, expedited examinations will be arranged, authoritative experts will be invited to participate in consultations, and the relevant data of the progression stagnation and subsequent measures will be recorded in detail.

Images