Cardiovascular nursing medical resource allocation system

Abstract

The invention discloses a cardiovascular nursing medical resource allocation system, and relates to the technical field of medical information, and the system comprises a multi-sensor data collection network which is used for collecting various signal data related to cardiovascular health from a patient and an environment; and the data integration and preprocessing module is used for performing format unification, noise suppression and feature extraction on the original data acquired by the multiple sensors. According to the cardiovascular nursing medical resource allocation system, real-time monitoring and accurate early warning of the cardiovascular state are realized through accurate data acquisition, data preprocessing and mixed deep learning risk assessment technologies; meanwhile, hospital resources are dynamically dispatched through multi-objective optimization and a genetic algorithm, and it is ensured that optimal treatment resources can be rapidly allocated when acute events occur.

Description

Technical Field

[0001] The present invention relates to the field of medical information technology, and in particular to a cardiovascular care medical resource allocation system. Background Art

[0002] In recent years, cardiovascular diseases have shown a high incidence trend worldwide and have become one of the important factors threatening human life and health. Traditional cardiovascular care mainly relies on regular monitoring and manual judgment, and has the following shortcomings: data collection and processing are scattered, and existing monitoring equipment often works independently and cannot form a unified data platform, which makes it difficult to conduct a comprehensive and dynamic assessment of the patient's cardiovascular status; abnormal event predictions are not timely enough. Due to the limitations of manual analysis and simple statistical methods, the early warning capabilities of acute cardiovascular events (such as arrhythmias and myocardial ischemia) are insufficient, and the best time for treatment is easily missed; resource allocation response is delayed. When facing sudden acute events, the dispatch of medical resources (including emergency equipment and professionals) often cannot reach the patient's location in the shortest time, reducing the efficiency of rescue; information interaction and remote collaboration are insufficient, and it is difficult for medical staff to obtain comprehensive monitoring data and early warning information in real time, and it is also difficult to communicate with remote experts in a timely manner, affecting the formulation and implementation of the overall treatment plan. Summary of the invention

[0003] 1. Technical issues to be resolved

[0004] In view of the shortcomings of the existing technology, the present invention provides a cardiovascular care medical resource allocation system, which solves the problems in the existing technology of discontinuous data collection, insufficient signal processing, low level of intelligent risk assessment, non-dynamic resource scheduling and insufficient data security protection.

[0005] (II) Technical solution

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a cardiovascular care medical resource allocation system, the system comprising:

[0007] A multi-sensor data acquisition network for collecting multiple signal data related to cardiovascular health from patients and the environment;

[0008] A data integration and preprocessing module is used to unify the format, suppress noise and extract features of the raw data collected by multiple sensors. The preprocessing includes using Kalman filtering and wavelet transform technology to obtain high-quality data for subsequent analysis;

[0009] The evaluation and early warning module uses a hybrid deep learning model and statistical analysis method to perform real-time anomaly detection and status evaluation on preprocessed data, thereby predicting the risk of acute cardiovascular events in patients and outputting early warning signals;

[0010] The resource scheduling module automatically calculates a resource scheduling plan based on the risk level provided by the early warning unit and the current hospital resource inventory using a multi-objective optimization strategy based on a genetic algorithm. This scheduling plan achieves the best allocation while taking into account response speed, treatment effect and balanced resource utilization;

[0011] The information interaction module is used to display real-time monitoring data, early warning information, and resource scheduling results to medical staff in a graphical interface, and supports remote diagnosis and treatment instructions and feedback to achieve data sharing and collaborative treatment;

[0012] The data transmission module ensures the security and integrity of data exchange between the above modules through encryption protocols.

[0013] Preferably, the multiple signal data include electrocardiogram, blood pressure, blood oxygen and medical equipment status information, and the multi-sensor data acquisition network includes an electrocardiogram sensor, a blood pressure monitor, a blood oxygen sensor and a medical equipment status acquisition device; each acquisition device performs real-time data acquisition for different physiological or equipment parameters.

[0014] Preferably, the data integration and preprocessing module performs unified format conversion, noise suppression and feature extraction on the raw data from each acquisition device; the preprocessing process includes:

[0015] (a1) Using the Kalman filter algorithm, the implementation steps are as follows:

[0016] (a11) The predicted state is calculated based on the state transfer matrix F and the observation matrix H. The formula for calculating the predicted state is: , where F can be set as the unit matrix or obtained based on the system dynamic modeling; the observation matrix H is determined by the sensor characteristics;

[0017] (a12) Calculate the prediction error covariance, the calculation formula is:

[0018] ; Where F is the state transfer matrix, Q is the process noise covariance, and its value can be determined according to the sensor accuracy and environmental noise;

[0019] (a13) Using the formula:

[0020] ; Calculate the Kalman gain, where H is the observation matrix and R is the measurement noise covariance, which is obtained through statistics of actual collected data;

[0021] (a14) Update the state using the observed value z(k), and the calculation formula is:

[0022] ; (a15) Through the formula:

[0023] ; Update error covariance; (a2) The continuous signal S(t) is decomposed by wavelet transform and the multi-scale coefficient C(j,k) is calculated. The calculation formula is:

[0024] ;

[0025] in, is the function after the mother wavelet function is transformed by scale j and translation k; multi-scale feature extraction is performed on continuous signals to ensure the accuracy of subsequent analysis data; among them, The calculation formula is:

[0026] ;

[0027] Among them, each scale j and translation k is determined by the design requirements and is used to extract local features in the signal.

[0028] Preferably, the assessment and warning module adopts a hybrid deep learning model, which includes a convolutional neural network and a long short-term memory network. The convolutional neural network is used to capture local time domain features, and the long short-term memory network is used to analyze long-term trends. The two are combined to perform real-time identification and risk prediction of patients' abnormal heart rate and arrhythmia, and output warning signals based on statistical confidence intervals.

[0029] Preferably, the implementation steps of the hybrid deep learning model in the evaluation and early warning module are:

[0030] (b1) Use convolutional neural network to extract local features of preprocessed data. The convolution operation formula is: ;

[0031] Where w(j) is the convolution kernel parameter and b is the bias; (b2) The output of the convolution layer is passed to the long short-term memory network. The long short-term memory network uses the formula to calculate the internal state. The calculation formula is:

[0032] ;

[0033] Among them, f t is the forget gate coefficient, i t is the input gate coefficient, W and U are weight matrices, and b′ is the bias; (b3) The LSTM output is mapped to m risk feature scores p through the fully connected layer ᵢ (i=1,2,…,m), and the risk index R is calculated using the following formula: ;

[0034] in, is the preset weight coefficient; (b4) When the risk index R is greater than the set threshold R threshold (R threshold =0.8), the unit outputs an acute cardiovascular event warning signal.

[0035] Preferably, the resource scheduling module realizes optimal resource scheduling by constructing a multi-objective optimization model based on the acute event risk data output by the early warning unit and the hospital resource inventory information. The optimization model uses a genetic algorithm for global search, and the objective function simultaneously considers minimization of response time, maximization of treatment benefits, and balanced resource distribution.

[0036] Preferably, the resource scheduling module automatically calculates the resource scheduling plan based on the risk index R output by the early warning unit and the current resource inventory data of the hospital, including the following steps: (c1) receiving the risk index R and the hospital resource inventory data from the evaluation and early warning module, wherein the inventory data items are the number of emergency beds B, the number of medical equipment M, and the number of specialist doctors D; (c2) constructing the objective function F, which is defined as: F=β1·T response -β2·E treatment +β3·U resource ;

[0037] Where: T response Indicates the warning response time; E treatment represents the expected treatment benefit, which is calculated through historical statistical data; U resource It represents the resource utilization balance index, which is calculated based on the allocation ratio of each resource; among them, T response The calculation formula is:

[0038] T response =t pre -t now ; where t pre is the warning trigger time, t now is the current time; (c3) Setting constraints, the process includes: (c31) For each resource i, there is an allocation x i Satisfies 0≤x i ≤Inventory i , where x i Indicates the allocation amount, Inventory i represents the inventory; (c32) Each resource allocation must meet the minimum demand N i , that is, x i ≥N i ; (c4) Use genetic algorithm for global optimization.

[0039] Preferably, the global optimization using a genetic algorithm comprises the following steps:

[0040] (c41) Generate an initial population, each individual is a set of resource allocation solutions {x1,x2,…,x k}; (c42) Evaluate individuals according to the fitness function f(x)=–F (the fitness value is inversely proportional to F), where F is the objective function value; (c43) Select according to the roulette method, and use single-point or multi-point crossover and mutation operators to generate a new generation of solutions; (c44) Repeat (c42) to (c43) until the preset number of iterations is reached or the fitness converges, and output the optimal scheduling plan.

[0041] Preferably, the information interaction module includes a touch screen display module, a wireless data transmission module and a remote communication interface, the touch screen display module is used to display the patient status and resource scheduling plan in real time, and the wireless data transmission module supports remote medical experts to review and intervene in the scheduling plan.

[0042] Preferably, the data transmission module adopts the SSL / TLS encryption protocol when transmitting data between the modules of the system, and the implementation steps include: (d1) before the communication session starts, the symmetric encryption key is encrypted using the RSA algorithm; (d2) during the actual data transmission process, the AES-256 algorithm is used for symmetric encryption to ensure the security of the ciphertext; (d3) after each communication is established, a summary of the transmitted data is generated by the SHA-256 hash function, and the summary is verified at the receiving end to ensure that the data has not been tampered with and to verify the data integrity.

[0043] (III) Beneficial effects

[0044] The present invention provides a cardiovascular care medical resource allocation system. It has the following beneficial effects:

[0045] (I) The cardiovascular care medical resource allocation system achieves real-time monitoring and accurate early warning of cardiovascular status through precise data collection, data preprocessing and hybrid deep learning risk assessment technology. At the same time, it uses multi-objective optimization and genetic algorithms to dynamically schedule hospital resources to ensure that the best treatment resources can be quickly allocated when acute events occur. In addition, the system also ensures the collaborative work and data transmission security between modules through efficient information interaction and strict data security measures.

[0046] (ii) The cardiovascular care medical resource allocation system, this technical solution significantly shortens the response time of acute cardiovascular events, reduces the misdiagnosis and missed diagnosis rates, improves the patient treatment success rate and the overall medical resource utilization efficiency, and at the same time has good scalability and adaptability, providing an efficient, reliable and safe technical means for clinical emergency and cardiovascular care.

[0047] (III) The cardiovascular care medical resource allocation system integrates multi-sensor data acquisition, data preprocessing, hybrid deep learning risk assessment, dynamic resource optimization scheduling based on genetic algorithms, and security information interaction. The system achieves accurate monitoring and prediction of the patient's cardiovascular status through efficient collaboration among various modules, and automatically adjusts the allocation of medical resources according to real-time risk conditions, thereby improving the response speed to acute events and the overall efficiency of treatment, and effectively improving the quality of patient treatment and the level of safety assurance. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a schematic diagram of the overall framework of the present invention;

[0049] Figure 2 This is a schematic diagram of the process of using genetic algorithm for global optimization in the present invention. DETAILED DESCRIPTION

[0050] 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.

[0051] See also Figure 1 and Figure 2 The present invention provides a technical solution: a cardiovascular care medical resource allocation system, the system comprising:

[0052] A multi-sensor data acquisition network is used to collect various signal data related to cardiovascular health from patients and the environment. In other words, the multi-sensor data acquisition network is responsible for collecting signal data from patients and the environment;

[0053] Data integration and preprocessing module, which is used to unify the format, suppress noise and extract features of the raw data collected by multiple sensors. Preprocessing includes the use of Kalman filtering and wavelet transform technology to obtain high-quality data for subsequent analysis;

[0054] The evaluation and early warning module uses a hybrid deep learning model and statistical analysis methods to perform real-time anomaly detection and status evaluation on preprocessed data, thereby predicting the risk of acute cardiovascular events in patients and outputting early warning signals; that is, the evaluation and early warning module uses deep learning and statistical methods to identify anomalies in real time and calculate risk indexes;

[0055] The resource scheduling module automatically calculates a resource scheduling plan based on the risk level provided by the early warning unit and the current hospital resource inventory using a multi-objective optimization strategy based on a genetic algorithm. This scheduling plan achieves the best allocation while taking into account response speed, treatment effect, and balanced resource utilization. In other words, the resource scheduling module automatically calculates the best scheduling plan based on the risk index and hospital inventory data by constructing optimization objectives and constraints.

[0056] The information interaction module is used to display real-time monitoring data, early warning information and resource scheduling results to medical staff in a graphical interface, and supports remote diagnosis and treatment instructions and feedback to achieve data sharing and collaborative treatment; that is, the information interaction module displays various types of information through a graphical interface and supports remote interaction;

[0057] The data transmission module ensures the security and integrity of data exchange between the above modules through encryption protocols, that is, the data transmission module uses encryption technology to achieve secure transmission of data between modules.

[0058] The various signal data include electrocardiogram, blood pressure, blood oxygen and medical equipment status information. The multi-sensor data acquisition network includes electrocardiogram sensor, blood pressure monitor, blood oxygen sensor and medical equipment status acquisition device; each acquisition device performs real-time data acquisition for different physiological or equipment parameters;

[0059] It should be further explained that, in the specific implementation process, the electrocardiogram sensor collects the patient's electrocardiogram signal at a sampling frequency of 250Hz, and has a built-in high-precision ADC module. The analog signal is converted into digital waveform data through an analog-to-digital converter (ADC), and the data format is 16-bit binary; the blood pressure monitor collects the patient's systolic and diastolic blood pressure data at 1 time / second, and outputs a numerical signal with a signal range of 0 to 300 mmHg. The blood oxygen sensor collects the patient's blood oxygen saturation data at 2 times / second. The blood oxygen sensor collects data twice per second to measure the blood oxygen saturation, and the output range is 0% to 100%. The medical equipment status acquisition device: uses an RS-485 bus interface, and the data is packaged in a predetermined protocol format ISO / IEC11898-1 format, and the working status of each device is transmitted once per second. The working status includes the switch status and fault code.

[0060] The data integration and preprocessing module performs unified format conversion, noise suppression and feature extraction on the raw data from each acquisition device; the preprocessing process includes:

[0061] (a1) Using the Kalman filter algorithm, the implementation steps are as follows:

[0062] (a11) The predicted state is calculated based on the state transfer matrix F and the observation matrix H. The formula for calculating the predicted state is: ;

[0063] (a12) Calculate the prediction error covariance, the calculation formula is:

[0064] ; Where F is the state transfer matrix and Q is the process noise covariance;

[0065] (a13) Using the formula:

[0066] ; Calculate the Kalman gain, where H is the observation matrix and R is the measurement noise covariance;

[0067] (a14) According to the formula:

[0068] ; Update status; (a15) Through the formula:

[0069] ; Update error covariance; (a2) The continuous signal S(t) is decomposed by wavelet transform and the multi-scale coefficient C(j,k) is calculated. The calculation formula is:

[0070] ;

[0071] in, It is the function after the mother wavelet function is transformed by scale j and translation k; multi-scale feature extraction is performed on continuous signals to ensure the accuracy of subsequent analysis data.

[0072] The evaluation and early warning module uses a hybrid deep learning model, which feeds preprocessed high-quality data into the hybrid model. The model consists of two main parts:

[0073] Convolutional Neural Network (CNN): used to extract local time domain features, using multiple convolutional layers and pooling layers, and the convolution kernel parameters are optimized through back propagation;

[0074] Long Short-Term Memory Network (LSTM): used to capture long-term dependencies in time series data and solve the problem of insufficient short-term memory;

[0075] The two work together to achieve real-time identification of abnormal patterns such as abnormal heart rate and arrhythmia in patients, and output statistical confidence intervals as a basis for risk warning.

[0076] The implementation steps of the hybrid deep learning model in the evaluation and early warning module are:

[0077] (b1) Convolutional layer operation:

[0078] For the input data x, the output is calculated through the convolution operation; that is, the convolutional neural network is used to extract local features of the preprocessed data. The convolution operation formula is: ;

[0079] Where w(j) is the convolution kernel parameter, b is the bias, and n is the convolution window size; (b2) LSTM layer operation:

[0080] The output sequence of the convolutional layer is passed into the long short-term memory network. The long short-term memory network uses the formula to calculate the internal state. The calculation formula is:

[0081] ;

[0082] Among them, f t is the forget gate coefficient, i t is the input gate coefficient, W, U are weight matrices, b′ is the bias, and f t and i t They are calculated by sigmoid function, W, U and b′ are model parameters; (b3) Full connection mapping and risk index calculation:

[0083] Map the LSTM output to m risk feature scores p through the fully connected layer ᵢ (i=1,2,…,m), and the risk index R is calculated using the following formula: ;

[0084] in, is the preset weight coefficient obtained through offline data training; (b4) Early warning judgment:

[0085] When the calculated risk index R is greater than the set threshold R threshold When, R threshold =0.8, the system outputs an acute cardiovascular event warning signal and triggers the subsequent resource scheduling process.

[0086] The resource scheduling module achieves optimal resource scheduling by constructing a multi-objective optimization model based on the acute event risk data and hospital resource inventory information output by the early warning unit. The optimization model uses a genetic algorithm for global search, and the objective function simultaneously considers minimization of response time, maximization of treatment benefits, and balanced resource distribution.

[0087] The resource scheduling module automatically calculates the resource scheduling plan based on the risk index R output by the early warning unit and the current resource inventory data of the hospital, including the following steps: (c1) receiving the risk index R and the hospital resource inventory data from the evaluation and early warning module, where the inventory data items are the number of emergency beds B, the number of medical devices M, and the number of specialists D; (c2) constructing the objective function F, which is defined as: F=β1·T response-β2·E treatment +β3·U resource ;

[0088] Where: T response Indicates the warning response time, which reflects the time difference from the triggering of the warning to the current time; E treatment represents the expected treatment benefit, which is calculated through historical statistical data; U resource It represents the resource utilization balance index, which is calculated based on the allocation ratio of each resource; β1, β2, β3 are adjustment coefficients, which are preset by the system administrator;

[0089] Among them, T response The calculation formula is:

[0090] T response =t pre -t now ; where t pre is the warning trigger time, t now is the current time, and the response delay is calculated; (c3) Setting constraints, the process includes: (c31) For each resource i, there is an allocation amount x i Satisfies 0≤x i ≤Inventory i , where x i Indicates the allocation amount, Inventory i represents the inventory; (c32) Each resource allocation must meet the minimum demand N i , that is, x i ≥N i；

[0091] The objective function constructed requires minimum response time, maximum treatment benefit and balanced resource utilization under the premise of satisfying the constraints; (c4) Genetic algorithm is used for global optimization.

[0092] The global optimization using genetic algorithm includes the following steps:

[0093] (c41) Generate several initial individuals according to the constraints, each of which represents a set of resource allocation schemes {x1, x2, …, x k}; (c42) Calculate the objective function value F for each individual, and define the fitness function as f(x)=–F to evaluate the individual, where the fitness value is inversely proportional to F, where F is the objective function value, and the higher the fitness value, the better the solution; (c43) Use the roulette method to select individuals with high fitness, and generate a new generation of solutions through single-point or multi-point crossover and random mutation; (c44) Repeat (c42) to (c43) until the preset number of iterations is reached or the fitness converges, and output the optimal scheduling solution.

[0094] The information interaction module includes a touch screen display module, a wireless data transmission module and a remote communication interface. The touch screen display module is used to display the patient status and resource scheduling plan in real time, and the wireless data transmission module supports remote medical experts to review and intervene in the scheduling plan;

[0095] It should be further explained that, in the specific implementation process, the touch screen display module uses a liquid crystal display with a resolution of not less than 1024×768 pixels to display the patient's real-time data, warning information and resource scheduling results in the form of charts; the wireless data transmission module transmits data packets according to the IEEE802.11 protocol, and attaches a CRC checksum to each data packet to ensure that there are no errors in the data packet transmission process; the remote communication interface supports two-way data interaction, allowing remote medical experts to approve the scheduling plan output by the system, and the approval information is encrypted and compared with the local data to ensure that the information is consistent before feedback to the system.

[0096] The data transmission module uses the SSL / TLS encryption protocol when transmitting data between various modules of the system. The implementation steps include: (d1) Before the communication session starts, the RSA algorithm is used to encrypt the symmetric encryption key to ensure the security of key exchange; (d2) During the actual data transmission process, the AES-256 algorithm is used to symmetric encrypt all transmitted data to ensure the security of the ciphertext; (d3) After each communication is established, the system uses the SHA-256 algorithm to calculate the data summary and compare it at the receiving end to verify that the data has not been tampered with during the transmission process.

[0097] It should be further explained that, in the specific implementation process, during the resource dispatch of the hospital's cardiovascular intensive care unit, the electrocardiogram sensor, blood pressure monitor and blood oxygen sensor installed on the patient's bed collect data at sampling frequencies of 250Hz, 1Hz and 2Hz respectively, and transmit various medical equipment status data through the RS-485 interface; the system sends these raw data to the data integration and preprocessing module, first uses the Kalman filter algorithm to suppress noise on the continuous signal, and uses wavelet transform to extract multiple scale features of the electrocardiogram signal, thereby forming high-quality feature data input;

[0098] The preprocessed data first enters the CNN layer, extracts local time domain features through multiple convolution kernels, and then enters the LSTM layer to analyze time series changes; the fully connected layer maps the LSTM output to m risk feature scores and calculates the risk index R; when the calculated value of a patient's risk index reaches 0.85 (greater than the preset threshold of 0.8), the system automatically issues an acute cardiovascular event warning signal and records the warning trigger time t pre;After receiving the warning signal, the system immediately calls the resource scheduling module to obtain the inventory data of emergency beds B, spare medical equipment M and specialist doctors D in the current ICU;

[0099] According to the formula: F = β1·T response -β2·E treatment +β3·U resource ; Calculate the warning response time T respnse =t pre −t now As well as the expected treatment benefit and resource utilization balance indicators; the genetic algorithm generates the initial population, evaluates the fitness, and outputs the optimal resource allocation plan after multiple iterations, including allocating 1 emergency bed, 1 emergency monitor and 2 specialists to quickly rush to the patient's bedside; the system displays the patient's physiological data, early warning information and scheduling plan in real time on the touch screen, and sends the data to the remote consultation center through the wireless transmission module, and transmits it through RSA and AES-256 encrypted transmission, and uses SHA-256 verification to ensure data integrity; remote experts confirm the scheduling plan through the interface, and the system records the approval results after data feedback, and finally completes the acute event handling;

[0100] During the implementation process, the system can complete data preprocessing, risk assessment and resource scheduling within a few seconds, greatly improving the response speed of acute cardiovascular events and ensuring the timeliness of patient rescue and treatment effectiveness.

[0101] It should be further explained that, in the specific implementation process, for cardiovascular monitoring and resource dispatch in mobile emergency vehicles, in the city emergency center, a mobile emergency vehicle equipped with this system is used to monitor the cardiovascular data of patients carried on the vehicle and patients passing by the emergency scene in real time, quickly determine whether there is an acute event, and make preliminary dispatch based on the limited medical resources in the vehicle, and at the same time send early warning information to nearby hospitals through the wireless network to coordinate subsequent resource support; the process is as follows:

[0102] The mobile emergency vehicle is equipped with a portable electrocardiogram sensor, a blood pressure monitor and a blood oximeter to collect patient data in real time at a standard sampling frequency; the vehicle's embedded small processing unit uses Kalman filtering and wavelet transform to pre-process the collected data, and the processed data is formatted and transmitted to the evaluation and warning module;

[0103] The preprocessed data is first processed by CNN to extract local features, and then the LSTM layer captures the patient's heart rate change trend and calculates the risk index R. When the risk index R reaches 0.82, the system issues an acute event warning signal and records the trigger time and location information.

[0104] Based on the existing resources in the vehicle: emergency medicines, portable defibrillators and resource inventory data of nearby hospitals, the system constructs an objective function through wireless network acquisition, and uses genetic algorithm optimization calculation to determine the best preliminary dispatch plan: start emergency rescue measures in the vehicle, and at the same time send a resource dispatch request to the nearest hospital, suggesting that the hospital reserve emergency beds and dispatch cardiovascular specialists;

[0105] The vehicle-mounted touch screen displays the patient's status, warning information and dispatch plan in real time; the data is transmitted to the emergency center via the IEEE802.11 wireless module, and all transmissions are encrypted with RSA / AES-256 and verified with SHA-256 to ensure information security; after the remote consultation platform receives the data, the expert confirms the dispatch plan in real time and returns the feedback information to the vehicle to further coordinate on-site emergency treatment;

[0106] The implementation process proved that the system also has efficient and accurate risk warning and resource dispatch capabilities in mobile emergency environments; through rapid judgment and coordination, the system significantly shortened the emergency response time and effectively connected on-site emergency treatment with follow-up hospital treatment, ensuring that patients receive the best medical resource support during the golden treatment period;

[0107] Through the above detailed technical solution description and two representative embodiments, the present invention discloses a specific implementation method of a cardiovascular care medical resource allocation system; the Kalman filter, wavelet transform, CNN+LSTM hybrid deep learning, multi-objective optimization based on genetic algorithm, and multiple encrypted data transmission technologies used in each module are disclosed in clear calculation formulas and steps, which not only ensures the close cooperation of various functional modules of the system, but also narrows the protection scope through specific implementation details, and reduces the review risk caused by vague description; the above embodiments fully demonstrate the actual application effect of the technical solution in hospital intensive care and mobile emergency, and reflect the creativity and practicality of the present invention in real-time warning and dynamic resource scheduling;

[0108] Each functional module adopts a layered structure design with standardized interfaces, which facilitates system expansion and subsequent upgrades. It can also be flexibly configured according to different hospitals or emergency scenarios. In addition to being suitable for hospital intensive care environments, this system can also be deployed in mobile emergency vehicles, emergency centers and other scenarios to achieve multi-level and multi-scenario cardiovascular care and resource scheduling, and effectively promote its application.

[0109] Through precise data collection, data preprocessing and hybrid deep learning risk assessment technology, real-time monitoring and accurate early warning of cardiovascular status are achieved; at the same time, multi-objective optimization and genetic algorithms are used to dynamically dispatch hospital resources to ensure that the best treatment resources can be quickly allocated when acute events occur; in addition, the system also ensures the collaborative work and data transmission security among modules through efficient information interaction and strict data security measures;

[0110] This technical solution significantly shortens the response time of acute cardiovascular events, reduces the misdiagnosis and missed diagnosis rates, improves the success rate of patient treatment and the overall efficiency of medical resource utilization, and at the same time has good scalability and adaptability, providing an efficient, reliable and safe technical means for clinical emergency and cardiovascular care.

[0111] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0112] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A cardiovascular care medical resource allocation system, characterized in that: The system includes: A multi-sensor data acquisition network for collecting multiple signal data related to cardiovascular health from patients and the environment; A data integration and preprocessing module is used to unify the format, suppress noise and extract features of the raw data collected by multiple sensors. The preprocessing includes using Kalman filtering and wavelet transform technology to obtain high-quality data for subsequent analysis; The evaluation and early warning module uses a hybrid deep learning model and statistical analysis method to perform real-time anomaly detection and status evaluation on preprocessed data, thereby predicting the risk of acute cardiovascular events in patients and outputting early warning signals; The resource scheduling module automatically calculates a resource scheduling plan based on the risk level provided by the early warning unit and the current hospital resource inventory using a multi-objective optimization strategy based on a genetic algorithm; The information interaction module is used to display real-time monitoring data, early warning information, and resource scheduling results to medical staff in a graphical interface, and supports remote diagnosis and treatment instructions and feedback to achieve data sharing and collaborative treatment; The data transmission module ensures the security and integrity of data exchange between modules through encryption protocols.

2. A cardiovascular care medical resource allocation system according to claim 1, characterized in that: The multiple signal data include electrocardiogram, blood pressure, blood oxygen and medical equipment status information, and the multi-sensor data acquisition network includes an electrocardiogram sensor, a blood pressure monitor, a blood oxygen sensor and a medical equipment status acquisition device; each acquisition device performs real-time data acquisition for different physiological or equipment parameters.

3. A cardiovascular care medical resource allocation system according to claim 2, characterized in that: The data integration and preprocessing module performs unified format conversion, noise suppression and feature extraction on the raw data from each acquisition device; The pretreatment process includes: (a1) Using the Kalman filter algorithm, the implementation steps are as follows: (a11) Calculate the predicted state according to the state transfer matrix F and the observation matrix H; (a12) Calculate the prediction error covariance; (a13) calculating the Kalman gain; (a14) Update status; (a15) Update error covariance; (a2) The continuous signal S(t) is decomposed using wavelet transform and the multi-scale coefficients are calculated.

4. A cardiovascular care medical resource allocation system according to claim 3, characterized in that: The assessment and early warning module adopts a hybrid deep learning model, which includes a convolutional neural network and a long short-term memory network. The convolutional neural network is used to capture local time domain features, and the long short-term memory network is used to analyze long-term trends. The two are combined to perform real-time identification and risk prediction of patients' abnormal heart rate and arrhythmia, and output early warning signals based on statistical confidence intervals.

5. A cardiovascular care medical resource allocation system according to claim 4, characterized in that: The implementation steps of the hybrid deep learning model in the evaluation and early warning module are: (b1) Use convolutional neural network to extract local features of preprocessed data; (b2) Pass the output of convolutional layer to long short-term memory network, which uses formula to calculate internal state; (b3) Map LSTM output to m risk feature scores p through fully connected layer ᵢ (i=1,2,…,m), and calculate the risk index R; (b4) When the risk index R is greater than the set threshold R threshold When the unit outputs an acute cardiovascular event warning signal.

6. A cardiovascular care medical resource allocation system according to claim 5, characterized in that: The resource scheduling module achieves optimal resource scheduling by constructing a multi-objective optimization model based on the acute event risk data output by the early warning unit and the hospital resource inventory information. The optimization model uses a genetic algorithm for global search, and the objective function simultaneously considers minimization of response time, maximization of treatment benefits, and balanced resource distribution.

7. A cardiovascular care medical resource allocation system according to claim 6, characterized in that: The resource scheduling module automatically calculates the resource scheduling plan based on the risk index R output by the early warning unit and the current resource inventory data of the hospital, including the following steps: (c1) receiving the risk index R and the hospital resource inventory data from the evaluation and early warning module, where the inventory data items are the number of emergency beds B, the number of medical equipment M, and the number of specialist doctors D; (c2) constructing the objective function F, which is defined as: F=β1·T response -β2·E treatment +β3·U resource ; Where: T response Indicates the warning response time; E treatment represents the expected treatment benefit; U resource represents the resource utilization balance index; where T response The calculation formula is: T response =t pre -t now ; where t pre is the warning trigger time, t now is the current time; (c3) Setting constraints, the process includes: (c31) For each resource i, there is an allocation x i Satisfies 0≤x i ≤Inventory i , where x i Indicates the allocation amount, Inventory i represents the inventory; (c32) each resource allocation satisfies the minimum demand N i , that is, x i ≥N i ; (c4) Use genetic algorithm for global optimization.

8. A cardiovascular care medical resource allocation system according to claim 7, characterized in that: The global optimization using genetic algorithm comprises the following steps: (c41) Generate an initial population, each individual is a set of resource allocation solutions {x1,x2,…,x k }; (c42) Evaluate the individuals according to the fitness function f(x)=–F, where F is the objective function value; (c43) Select according to the roulette method, and use single-point or multi-point crossover and mutation operators to generate a new generation of solutions; (c44) Repeat (c42) to (c43) until the preset number of iterations is reached or the fitness converges, and output the optimal scheduling plan.

9. A cardiovascular care medical resource allocation system according to claim 8, characterized in that: The information interaction module includes a touch screen display module, a wireless data transmission module and a remote communication interface. The touch screen display module is used to display the patient status and resource scheduling plan in real time, and the wireless data transmission module supports remote medical experts to review and intervene in the scheduling plan.

10. A cardiovascular care medical resource allocation system according to claim 9, characterized in that: The data transmission module adopts the SSL / TLS encryption protocol when transmitting data between the modules of the system, and the implementation steps include: (d1) before the communication session starts, the symmetric encryption key is encrypted using the RSA algorithm; (d2) during the actual data transmission process, the AES-256 algorithm is used for symmetric encryption; (d3) after each communication is established, a summary of the transmitted data is generated using the SHA-256 hash function and verified at the receiving end.

Images