A smart medical management system and method based on big data

Abstract

This invention relates to the field of healthcare information technology, and more particularly to a smart healthcare management system and method based on big data. The system includes: collecting and cleaning raw data; using an improved information gain ratio feature selection algorithm to select optimal features from the cleaned data; constructing a dynamic adaptive system; updating the state of the dynamic adaptive system based on the optimal features; utilizing complex network theory to treat the state of the dynamic adaptive system as nodes in a network and synchronizing them; processing the synchronized output results of the network nodes; and monitoring the patient's health status. This invention addresses the problems of low efficiency in medical data processing, inability to dynamically adapt to data changes, insufficient clinical decision support, inadequate data security and privacy protection, limited ability to handle multiple data sources and types, inability to reflect real-time changes in the patient's health status, and inability to effectively assist doctors in formulating treatment plans and health management strategies, thus affecting treatment outcomes.

Description

Technical Field

[0001] The present invention relates to the field of medical care information technology, and in particular to a big data-based smart medical management system and method. Background Art

[0002] In the field of modern healthcare, efficient processing and analysis of medical data is essential to optimizing patient treatment outcomes and management strategies. With the advancement of medical technology and the popularization of digital medical information, medical institutions have accumulated a large amount of medical data involving electronic health records, medical images, laboratory test results, and vital signs data from wearable devices. These medical data are of various types, including text, images, and time series, and each type has its own unique processing requirements and challenges.

[0003] The development and application of big data technology has made it possible to improve the depth and breadth of medical data processing, improve the accuracy of diagnosis, formulate personalized treatment plans, and more effectively manage health and prevent diseases through more refined data analysis methods. In addition, the progress of medical data analysis is also expected to play an important role in public health management, helping policymakers and medical institutions to better respond to health crises and optimize resource allocation through disease pattern recognition and health trend prediction.

[0004] The above-mentioned technologies have the following technical problems: inefficient medical data processing, inability to dynamically adapt to data changes, insufficient clinical decision support, inadequate data security and privacy protection, and limited ability to process multiple data sources and types, which affect prediction accuracy and decision quality, cannot reflect real-time changes in patients' health status or adapt to complex patterns in the disease process, cannot effectively assist doctors in formulating personalized treatment plans or adjusting treatment strategies in a timely manner, and affect the comprehensiveness and effectiveness of medical services. Summary of the invention

[0005] The present invention provides a smart medical management system and method based on big data to solve the problems of low efficiency in medical data processing, inability to dynamically adapt to data changes, insufficient clinical decision support, insufficient data security and privacy protection, and limited ability to process multiple data sources and types, which affect prediction accuracy and decision quality, cannot reflect real-time changes in patient health status or adapt to complex patterns in the course of the disease, cannot effectively assist doctors in formulating personalized treatment plans or adjusting treatment strategies in a timely manner, and affect the comprehensiveness and effectiveness of medical services.

[0006] The present invention provides a big data-based intelligent medical management system and method, which specifically includes the following technical solutions:

[0007] A smart medical management method based on big data, comprising the following steps:

[0008] S1. Collect raw data, clean the raw data, and obtain cleaned data; use the improved information gain ratio feature selection algorithm to select features from the cleaned data to obtain the optimal features;

[0009] S2. Construct a dynamic adaptive system and update the state of the dynamic adaptive system based on the optimal features. Use complex network theory to regard the state of the dynamic adaptive system as a node in the network, synchronize the network nodes, and obtain the synchronized output results of the network nodes. Use adaptive filters to process the synchronized output results of the network nodes to obtain the prediction results. Based on the prediction results, monitor the patient's health status.

[0010] Preferably, the S1 specifically includes:

[0011] In the process of data cleaning, the cleaning complexity is calculated. The specific formula is:

[0012]

[0013] in, Indicates the cleaning complexity; Indicates the original data The number of outliers; Indicates the number of types of outliers in the original data; Indicates the original data The amount of data associated with each outlier; is a positive number used to avoid zero denominator; and is the adjustment parameter; Indicates the original data The standard deviation of the outliers; and is a parameter used to adjust the strength of the adaptive factor.

[0014] Preferably, the S1 specifically includes:

[0015] Based on the improved information gain ratio feature selection algorithm, the information gain rate of each feature in the cleaned data is calculated, and the feature with the highest information gain rate is selected as the optimal feature; and the feature selection process is optimized based on the cleaning complexity.

[0016] Preferably, the S1 specifically includes:

[0017] The cleaning complexity is normalized and used as a weight factor in the improved information gain ratio feature selection algorithm to adjust the feature selection process.

[0018] Preferably, the S2 specifically includes:

[0019] Construct a dynamic adaptive system. Based on the optimal features and combined with the previous state of the dynamic adaptive system, update the state of the dynamic adaptive system. The state update formula of the dynamic adaptive system is:

[0020]

[0021] in, express Dynamic adaptive system state at each moment; yes Dynamic adaptive system state at each moment; is the decay factor that controls the degree to which the previous state of a dynamic adaptive system is preserved in the new state; and is a dynamic adaptive system parameter; represents the optimal feature.

[0022] Preferably, the S2 specifically includes:

[0023] The specific formula for synchronizing network nodes is:

[0024]

[0025] in, It's time The synchronous output results of the network nodes; It is Nodes at time The adjacency matrix of It is Nodes at time Status; and They are the hyperbolic sine function and the hyperbolic cosine function respectively; is the total number of network nodes.

[0026] Preferably, the S2 specifically includes:

[0027] Based on the synchronous output results of the network nodes, the response strength of the adaptive filter is dynamically adjusted, and the synchronous output results of the network nodes are processed by the adaptive filter to generate the final health trend prediction and disease pattern output as the prediction result.

[0028] A smart medical management system based on big data, including the following parts:

[0029] Data collection module, data cleaning module, feature selection module, dynamic adaptive module, state synchronization module, prediction module;

[0030] Data collection module, collects raw data; transmits raw data to data cleaning module;

[0031] The data cleaning module cleans the collected raw data to obtain the cleaned data and calculates the cleaning complexity; the cleaned data and the cleaning complexity are transmitted to the feature selection module;

[0032] The feature selection module uses an improved information gain ratio feature selection algorithm to select the optimal features based on the cleaning complexity of the cleaned data; and transmits the optimal features to the dynamic adaptive module;

[0033] The dynamic adaptive module builds a dynamic adaptive system, adjusts internal parameters based on the optimal features, updates the dynamic adaptive system state, and transmits the dynamic adaptive system state to the state synchronization module;

[0034] The state synchronization module uses complex network theory to regard the state of the dynamic adaptive system as a node in the network, synchronizes the state of the dynamic adaptive system, and obtains the synchronized output results of the network nodes; the synchronized output results of the network nodes are transmitted to the prediction module;

[0035] The prediction module processes the synchronous output results of the network nodes through an adaptive filter to generate the final output of health trend prediction and disease pattern as the prediction result; based on the prediction result, the patient's health status is monitored.

[0036] The beneficial effects of the technical solution of the present invention are:

[0037] 1. By introducing the data cleaning module, it is possible to effectively remove erroneous data, fill in missing values, and perform data normalization and desensitization of sensitive information. This not only improves the accuracy and availability of medical data, but also ensures the efficiency of data processing and the security of data privacy. Using the improved information gain ratio feature selection algorithm, the smart medical management system can more accurately select the features that have the greatest impact on health outcome prediction, thereby improving the prediction performance and the quality of related health management decisions.

[0038] 2. By building a dynamic adaptive system, it is possible to dynamically adjust internal parameters according to specific changes in medical data, so that the dynamic adaptive system can more accurately identify disease patterns and predict health trends. It is suitable for processing medical data in long-term health monitoring, such as continuous monitoring data of patients with chronic diseases. The self-adjustment of the dynamic adaptive system reduces the need for human intervention and improves the reliability and stability of the prediction results, enabling it to maintain efficient operation in a changing medical environment.

[0039] 3. Through real-time and accurate health trend predictions, doctors and medical providers can obtain key decision support information to better formulate treatment plans and health management strategies; the prediction results provided by the smart medical management system can help medical professionals make more scientific and personalized medical decisions in clinical practice, thereby improving treatment outcomes and patient satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a structural diagram of a smart medical management system based on big data described in the present invention;

[0041] Figure 2 This is a flow chart of a smart medical management method based on big data described in the present invention. DETAILED DESCRIPTION

[0042] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all 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.

[0043] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0044] The following is a detailed description of a specific solution of a smart medical management system and method based on big data provided by the present invention in conjunction with the accompanying drawings.

[0045] See attached Figure 1 , which shows a structure diagram of a smart medical management system based on big data provided by an embodiment of the present invention, the system includes the following parts:

[0046] Data collection module, data cleaning module, feature selection module, dynamic adaptive module, state synchronization module, prediction module;

[0047] a data collection module that collects raw data from a variety of medical data sources, including electronic health records, medical images, laboratory test results, and vital sign data from wearable devices; and transmits the raw data to a data cleaning module;

[0048] The data cleaning module cleans the collected raw data to obtain the cleaned data and calculates the cleaning complexity; the cleaned data and the cleaning complexity are transmitted to the feature selection module;

[0049] The feature selection module uses an improved information gain ratio feature selection algorithm to select the optimal features based on the cleaning complexity of the cleaned data; and transmits the optimal features to the dynamic adaptive module;

[0050] The dynamic adaptive module builds a dynamic adaptive system that can respond to and adapt to data changes, captures the time correlation and nonlinear characteristics of the optimal features, updates the state of the dynamic adaptive system, and transmits the state of the dynamic adaptive system to the state synchronization module;

[0051] The state synchronization module uses complex network theory to treat the state of the dynamic adaptive system as a node in the network. It analyzes health trends or disease patterns through the synchronization mechanism between network nodes, synchronizes the state of the dynamic adaptive system, and obtains the synchronized output results of the network nodes; the synchronized output results of the network nodes are transmitted to the prediction module;

[0052] The prediction module processes the synchronous output results of the network nodes through an adaptive filter to generate the final output of health trend prediction and disease pattern as the prediction result; the prediction result can be used for clinical decision support to help doctors make more accurate treatment and management decisions.

[0053] See attached Figure 2 , which shows a flow chart of a smart medical management method based on big data provided by an embodiment of the present invention, the method comprises the following steps:

[0054] S1. Collect raw data, clean the raw data, and obtain cleaned data; use the improved information gain ratio feature selection algorithm to select features from the cleaned data to obtain the optimal features;

[0055] The data collection module collects raw data from a variety of medical data sources, including electronic health records, medical images, laboratory test results, and vital signs data from wearable devices, etc., including various types of medical data such as text, images, and time series.

[0056] In order to improve data quality, the original data is cleaned to obtain cleaned data; data cleaning includes tasks such as removing erroneous data, filling missing values, data normalization, and desensitizing sensitive information. For example, for text data in electronic health records, natural language processing technology (NLP) is used to identify and eliminate irrelevant information; for time series data of vital signs, anomaly detection algorithms are used to eliminate obvious error values. Natural language processing technology and anomaly detection algorithms are both existing technologies and will not be described in detail here. During the data cleaning process, the cleaning complexity is also calculated. The specific calculation formula for cleaning complexity is:

[0057]

[0058] in, Indicates the cleaning complexity; Indicates the original data The number of outliers; Indicates the number of types of outliers in the original data; Indicates the original data The amount of data associated with each outlier; is a positive number added to avoid the denominator being zero; and is the adjustment parameter; Indicates the original data The standard deviation of outliers is used to measure the distribution stability of outliers in the original data; and is a parameter used to adjust the strength of the adaptive factor, through and The introduced adaptive factor can dynamically adjust the cleaning strategy according to the specific distribution of the original data and process the original data more accurately.

[0059] The cleaned data are integrated, and the improved information gain ratio feature selection algorithm is used to select the features that can most effectively predict health outcomes. By calculating the information gain and information gain rate of each feature in the cleaned data, the feature that contributes most to the classification task is selected as the optimal feature. It is regarded as an important indicator of data quality and is used to optimize the feature selection process. Specifically, the cleaning complexity Normalization processing is performed to obtain normalized cleaning complexity; the normalized cleaning complexity is used as a weight factor in the improved information gain ratio feature selection algorithm to adjust the feature selection process.

[0060] Cleaning complexity after normalization As a measure of data quality, the information gain ratio of each feature in the cleaned data is calculated. The specific formula is:

[0061]

[0062] in, Used to evaluate the information gain ratio of each feature in the cleaned data and preliminarily screen high-quality features in the cleaned data; It is the cleaned data; The cleaned data The total number of features in ; is the normalized cleaning complexity, which is used to measure the cleaned data The quality and credibility of each feature in is a positive number used to prevent the denominator from being zero; is the adjustment index, which controls the adjustment degree of the final information gain ratio; The cleaned data Information entropy of It is in the characteristics The cleaned data The conditional information entropy of Representation characteristics Information gain of It is a feature The split information is measured by the split information when using the feature After cleaning the data After segmentation, the sum of the uncertainties of the resulting subsets is the evaluation feature After cleaning the data An indicator of segmentation quality; smaller split information means that using the current feature The higher the purity of the data subset generated after segmentation, the more certain the classification result is. The calculation formula of split information is defined as:

[0063]

[0064] in, According to the characteristics After cleaning the data After segmentation, the The number of samples in the subset; The cleaned data The total number of samples in ; According to the characteristics After cleaning the data The number of different subsets generated after the split. The split information is used together with the information gain to obtain the "information gain rate". Based on the information gain rate, the optimal feature is selected; the specific calculation formula of the information gain rate is as follows:

[0065]

[0066] in, is the information gain rate, which is a classic feature selection criterion. It further selects the best features based on the high-quality features in the initially screened and cleaned data. Split the cleaned data When , information gain measures the reduction in data uncertainty; if a feature If there are many unique values ​​(such as user IDs), then based on the characteristics After cleaning the data The subset generated after segmentation may only contain a single data point, then the uncertainty (entropy) of the subset is reduced, thereby increasing the information gain, but features with low uncertainty are not helpful for generalization. Excessive split information means that the number of segmentations caused by the multi-valued nature of the feature exceeds the limit, resulting in overfitting of the prediction results. Selecting the feature with the highest information gain rate as the optimal feature and performing segmentation based on the optimal feature means that the current feature can provide the most effective information while maintaining the quality of data segmentation without unnecessarily increasing the complexity of the prediction results. The optimal feature is applied to subsequent health trend prediction and disease pattern recognition.

[0067] S2. Construct a dynamic adaptive system and update the state of the dynamic adaptive system based on the optimal features. Use complex network theory to regard the state of the dynamic adaptive system as a node in the network, synchronize the network nodes, and obtain the synchronized output results of the network nodes. Use adaptive filters to process the synchronized output results of the network nodes to obtain the prediction results. Based on the prediction results, monitor the patient's health status.

[0068] Taking into account the changes that may occur in medical data over time, such as the progression or regression of a patient's condition and the addition of new medical information, the dynamic adaptive module builds a dynamic adaptive system that can respond and adapt to changes in medical data. The dynamic adaptive system adjusts internal parameters based on the optimal features to more accurately predict and identify health trends and disease patterns.

[0069] Based on prior knowledge, such as average health indicators, disease incidence, etc., an initial state is set for the dynamic adaptive system; once the initial state is set, the dynamic adaptive system can start its dynamic operation; each new data input (optimal feature ) will lead to the update of the state of the dynamic adaptive system; the current state of the dynamic adaptive system Depends on its previous state and the current input (optimal feature ) The state update formula of the dynamic adaptive system is:

[0070]

[0071] in, Indicates the next moment (i.e. The dynamic adaptive system state at time ( is the current moment (i.e. The dynamic adaptive system state at time ( is a decay factor that controls the previous state of the dynamic adaptive system (i.e. state at that moment) in the new state (i.e. The degree of retention in the state at the moment, The closer it is to 1, the greater the influence of the previous state, and the slower the dynamic adaptive system responds to new information; is the optimal feature based on the current input The new information calculated, where and is a dynamic adaptive system parameter, weight and bias It is not fixed, but based on the optimal features of the input Dynamically adjust to changes in

[0072] By introducing a dynamic adaptive system, we can not only capture the time correlation and nonlinear characteristics of the optimal features, but also adjust the internal structure of the dynamic adaptive system in real time to best reflect and predict the changing trend of the optimal features. This is especially important for medical data analysis, because the health status of patients and related medical indicators may change in complex ways over time.

[0073] The state synchronization module uses complex network theory to treat the state of the dynamic adaptive system as nodes in the network. The nodes interact with each other through synchronization mechanisms, which is particularly important for medical data analysis because various factors in medical data are often interrelated and jointly affect health outcomes. The state synchronization of network nodes is used to improve the overall stability and predictive ability of the dynamic adaptive system:

[0074]

[0075] in, It's time The synchronous output result of the network nodes, that is, each node in the network is in time The overall status of It is Nodes at time The adjacency matrix represents the connection strength between different nodes and is used to adjust the interaction strength between nodes; It is Nodes at time Status; and They are the hyperbolic sine function and the hyperbolic cosine function, which are used to limit the range of network output; is the total number of network nodes.

[0076] The prediction module processes the synchronous output results of the network nodes through an adaptive filter to generate the final health trend prediction and disease pattern output, that is, the prediction result. The adaptive filter is used to smooth the synchronous output results of the network nodes, and the response strength of the adaptive filter is adjusted according to the synchronous output results of the network nodes, so as to ensure the accuracy and stability of the prediction results. The prediction result calculation formula of the adaptive filter is:

[0077]

[0078] in, Indicates at time The prediction results at time Indicates at time The prediction result at that time is the historical prediction; is a smoothing factor that balances historical forecasts and the current moment (i.e. moment) to ensure the smooth transition of the prediction results; It is an adaptive function that adjusts the response strength of the adaptive filter according to the complexity of the synchronous output results of the network nodes. It provides a prediction of the patient's future health status, which can be directly used for clinical decision support, such as predicting disease worsening, health improvement or stability, etc. Doctors can adjust treatment plans or conduct further diagnostic tests based on the prediction results.

[0079] In actual medical environments, by integrating the present invention into a smart medical management system, medical providers can use real-time updates of prediction results to continuously monitor the health status of patients. For example, in chronic disease management, by continuously tracking Doctors can timely understand the progress of the disease and adjust the treatment plan in time to achieve precision medicine and personalized treatment. In addition, the present invention can also be used in the field of public health to predict the trend of disease outbreaks by analyzing large-scale health data and provide data support for public health decision-making.

[0080] In summary, a smart medical management system and method based on big data has been completed.

[0081] The order of the embodiments of the invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0082] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0083] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A smart medical management method based on big data, characterized in that: The following steps are involved: S1. Collect the original data, clean the original data, and obtain the cleaned data; in the process of data cleaning, calculate the cleaning complexity, the specific formula is: ； in, Indicates the cleaning complexity; Indicates the original data The number of outliers; Indicates the number of types of outliers in the original data; Indicates the original data The amount of data associated with each outlier; is a positive number used to avoid zero denominator; and is the adjustment parameter; Indicates the original data The standard deviation of the outliers; and is a parameter used to adjust the strength of the adaptive factor; Based on the improved information gain ratio feature selection algorithm, the information gain ratio of each feature in the cleaned data is calculated, and the feature with the highest information gain ratio is selected as the optimal feature; and the feature selection process is optimized based on the cleaning complexity; S2. Build a dynamic adaptive system, based on the optimal features, combined with the previous state of the dynamic adaptive system, update the state of the dynamic adaptive system; the state update formula of the dynamic adaptive system is: ； in, express Dynamic adaptive system state at each moment; yes Dynamic adaptive system state at each moment; is the decay factor that controls the degree to which the previous state of a dynamic adaptive system is preserved in the new state; and is a dynamic adaptive system parameter; represents the optimal feature; Using complex network theory, the dynamic adaptive system state is regarded as a node in the network, and the network nodes are synchronized to obtain the synchronized output results of the network nodes; the specific formula for synchronizing network nodes is: ； in, It's time The synchronous output results of the network nodes; It is Nodes at time The adjacency matrix of It is Nodes at time Status; and They are the hyperbolic sine function and the hyperbolic cosine function respectively; is the total number of network nodes; The synchronous output results of the network nodes are processed by an adaptive filter to obtain a prediction result; based on the prediction result, the health status of the patient is monitored.

2. According to the big data-based smart medical management method of claim 1, it is characterized in that: The S1 specifically includes: The cleaning complexity is normalized and used as a weight factor in the improved information gain ratio feature selection algorithm to adjust the feature selection process.

3. The intelligent medical management method based on big data according to claim 1, characterized in that: The S2 specifically includes: Based on the synchronous output results of the network nodes, the response strength of the adaptive filter is dynamically adjusted, and the synchronous output results of the network nodes are processed by the adaptive filter to generate the final health trend prediction and disease pattern output as the prediction result.

4. A smart medical management system based on big data, applied to a smart medical management method based on big data as claimed in claim 1, characterized in that: Includes the following parts: Data collection module, data cleaning module, feature selection module, dynamic adaptive module, state synchronization module, prediction module; Data collection module, collects raw data; transmits raw data to data cleaning module; The data cleaning module cleans the collected raw data to obtain the cleaned data and calculates the cleaning complexity; the cleaned data and the cleaning complexity are transmitted to the feature selection module; The feature selection module uses an improved information gain ratio feature selection algorithm to select the optimal features based on the cleaning complexity of the cleaned data; and transmits the optimal features to the dynamic adaptive module; The dynamic adaptive module builds a dynamic adaptive system, adjusts internal parameters based on the optimal features, updates the dynamic adaptive system state, and transmits the dynamic adaptive system state to the state synchronization module; The state synchronization module uses complex network theory to regard the state of the dynamic adaptive system as a node in the network, synchronizes the state of the dynamic adaptive system, and obtains the synchronized output results of the network nodes; Transmit the synchronous output results of the network nodes to the prediction module; The prediction module processes the synchronous output results of the network nodes through an adaptive filter to generate the final output of health trend prediction and disease pattern as the prediction result; Based on the prediction results, the patient's health status is monitored.

Images