Real-time respirator parameter optimization system based on digital twinborn technology

Abstract

The invention relates to the technical field of breathing machines, in particular to a real-time breathing machine parameter optimization system based on a digital twinning technology, which comprises a patient modeling module for acquiring sensor data of an external sensor and establishing a patient digital model according to the sensor data; the breathing condition analysis module is used for analyzing the breathing machine data according to the patient digital model to obtain an analysis result; and the parameter adjusting module is used for adjusting the control parameters of the breathing machine according to the analysis result. Aiming at the problem of poor control effect of a respirator in the prior art, a patient modeling module is arranged in the scheme, a patient digital model is constructed through a digital twinning technology, and then a breathing condition analysis module analyzes respirator data according to the patient digital model to obtain a more accurate analysis result; and the subsequent parameter adjustment process is more in line with the actual situation of the user.

Description

Technical Field

[0001] The present invention relates to the technical field of ventilators, and particularly to a real-time ventilator parameter optimization system based on digital twin technology. Background Art

[0002] A ventilator is a medical device that can replace the function of autonomous ventilation and is widely used in patients with respiratory failure, during anesthesia respiratory management during surgery, respiratory support treatment, and first aid resuscitation processes, and can prevent and treat respiratory failure. In order to achieve a better respiratory replacement effect, corresponding sensors and control algorithms are usually designed in existing ventilators to realize the automatic adjustment process of various parameters of the ventilator.

[0003] For example, Chinese Patent CN201410072324.6 discloses a ventilator, whose working modes include a ventilation mode and an autonomous breathing mode. The ventilator includes: a respiratory rate monitor for monitoring the patient's respiratory rate and comparing it with a respiratory rate threshold; an end-expiratory positive pressure monitor for detecting the patient's end-expiratory positive pressure and comparing it with an end-expiratory positive pressure threshold; a rapid shallow breathing index monitor for detecting the patient's rapid shallow breathing index and comparing it with a rapid shallow breathing index threshold; an alarm device; and a controller. The input end of the controller is respectively connected to the respiratory rate monitor, the end-expiratory positive pressure monitor, and the rapid shallow breathing index monitor, and the output end of the controller is connected to the alarm device. Among them, in the autonomous breathing mode, when at least one of the respiratory rate, end-expiratory positive pressure, and rapid shallow breathing index exceeds the corresponding threshold, the controller controls the alarm device to give an alarm. It has a high degree of automation and does not require manual on-duty, which is beneficial to improving the medical staff efficiency.

[0004] However, in the actual implementation process, the inventor found that in such technical solutions, the corresponding control algorithms are usually calibrated for healthy people, and in the control process, the control algorithm is simply used to adjust various parameters to make the sensor readings meet the standard values. However, in the actual use process, the oxygen consumption of different types of patients is different, and the fixed control parameters cannot well meet these needs. Summary of the Invention

[0005] In view of the above problems existing in the prior art, a real-time ventilator parameter optimization system based on digital twin technology is provided.

[0006] The specific technical solution is as follows:

[0007] A real-time ventilator parameter optimization system based on digital twin technology includes:

[0008] A patient modeling module, the patient modeling module is connected to an external sensor and obtains sensor data of the external sensor, and the patient modeling module establishes a patient digital model according to the sensor data;

[0009] A respiratory condition analysis module, which is connected to a ventilator and collects ventilator data. The respiratory condition analysis module is connected to the patient modeling module, and the respiratory condition analysis module analyzes the ventilator data according to the patient digital model to obtain an analysis result;

[0010] A parameter adjustment module, which is connected to the respiratory condition analysis module and the ventilator. The parameter adjustment module adjusts the control parameters of the ventilator according to the analysis result.

[0011] On the other hand, the patient modeling module includes:

[0012] A data acquisition module, which extracts the sensor data from the external sensors;

[0013] A data processing module, which is connected to the data acquisition module. The data processing module preprocesses the sensor data to obtain preprocessed data;

[0014] A modeling module, which is connected to the data processing module. The modeling module generates the patient digital model according to the preprocessed data.

[0015] On the other hand, the patient modeling module includes:

[0016] A virtual model retrieval module, which determines the corresponding model type according to the patient basic information in the preprocessed data. The virtual model retrieval module determines the corresponding basic model according to the model type;

[0017] A model adjustment module, which is connected to the virtual model retrieval module. The model adjustment module adjusts the basic model according to the preprocessed data to generate the patient digital model.

[0018] On the other hand, the respiratory condition analysis module includes:

[0019] A ventilator interaction module, which is connected to the ventilator and collects the ventilator data;

[0020] A mapping table adjustment module, which is connected to the patient modeling module. The mapping table adjustment module adjusts the respiratory condition scoring table according to the patient digital model to obtain an adjusted table;

[0021] A condition comparison module, which is respectively connected to the ventilator interaction module and the mapping table adjustment module. The condition comparison module compares the ventilator data with the adjusted table to obtain the analysis result.

[0022] On the other hand, a plurality of ventilator sensors are respectively installed on the ventilator, and the ventilator sensors include a flow sensor, a pressure sensor, and an oxygen sensor;

[0023] The flow sensor is installed in the ventilator pipeline of the ventilator;

[0024] The pressure sensors are respectively installed at the outlet and the return port of the ventilator pipeline of the ventilator;

[0025] The oxygen sensor is installed in the oxygen mask of the ventilator.

[0026] On the other hand, the ventilator interaction module includes:

[0027] A data acquisition module, the data acquisition module is respectively connected to the ventilator sensors, and the data acquisition module acquires raw data from the ventilator sensors through corresponding data protocols;

[0028] A data format unification module, the data format unification module is connected to the data acquisition module, and the data format unification module converts the raw data according to the data protocol to form the ventilator data.

[0029] On the other hand, the mapping table adjustment module includes:

[0030] A model comparison module, the model comparison module obtains the patient digital model, and compares the patient digital model with the reference model to form model difference data;

[0031] A respiration prediction module, the respiration prediction module is connected to the model comparison module, and the respiration prediction module predicts respiration prediction data according to the model difference data;

[0032] A threshold calibration module, the threshold calibration module is connected to the respiration prediction module, and the threshold calibration module determines a respiration state threshold according to the respiration prediction data, and adjusts the respiration condition scoring table to obtain the adjusted table.

[0033] On the other hand, the condition comparison module includes:

[0034] A preliminary comparison module, the preliminary comparison module respectively compares each item in the ventilator data with the adjusted table to determine a preliminary comparison result;

[0035] A cause prediction module, the cause prediction module is connected to the preliminary comparison module, and the cause prediction module inputs into a decision tree model according to the preliminary comparison module to predict the analysis result.

[0036] On the other hand, the parameter adjustment module includes:

[0037] A parameter optimization module that inputs the analysis result into a parameter optimization model to obtain model output parameters;

[0038] The model output parameters include control parameters and sensor expected parameters;

[0039] A closed-loop adjustment module that is connected to the parameter optimization module. The closed-loop adjustment module controls the ventilator according to the control parameters and collects the ventilator data;

[0040] The closed-loop adjustment module also adjusts the control parameters according to the ventilator data until the ventilator data meets the sensor expected parameters.

[0041] The above technical solution has the following advantages or beneficial effects:

[0042] Aiming at the problem of poor control effect of the ventilator in the prior art, a patient modeling module is set up in this solution. A patient digital model is constructed through digital twin technology. Then, the breathing condition analysis module analyzes the ventilator data according to the patient digital model to obtain a more accurate analysis result, making the subsequent parameter adjustment process more in line with the actual situation of the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] With reference to the accompanying drawings, the embodiments of the present invention are described more fully. However, the accompanying drawings are only for illustration and explanation and do not constitute a limitation on the scope of the present invention.

[0044] Figure 1 It is a schematic diagram of the whole of the embodiment of the present invention;

[0045] Figure 2 It is a schematic diagram of the patient modeling module in the embodiment of the present invention;

[0046] Figure 3 It is a schematic diagram of the modeling module in the embodiment of the present invention;

[0047] Figure 4 It is a schematic diagram of the breathing condition analysis module in the embodiment of the present invention;

[0048] Figure 5 It is a schematic diagram of the ventilator interaction module in the embodiment of the present invention;

[0049] Figure 6 It is a schematic diagram of the mapping table adjustment module in the embodiment of the present invention;

[0050] Figure 7 It is a schematic diagram of the condition comparison module in the embodiment of the present invention. Detailed implementation mode

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work belong to the scope of protection of the present invention.

[0052] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments may be combined with each other.

[0053] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but it is not limited to the present invention.

[0054] The present invention includes:

[0055] A real-time ventilator parameter optimization system based on digital twin technology, as Figure 1 shown, includes:

[0056] A patient modeling module 1, the patient modeling module 1 is connected to an external sensor 1A and obtains sensor data of the external sensor 1A, and the patient modeling module 1 establishes a patient digital model according to the sensor data;

[0057] A breathing condition analysis module 2, the breathing condition analysis module 2 is connected to a ventilator 2A and collects ventilator data, the breathing condition analysis module 2 is connected to the patient modeling module 1, and the breathing condition analysis module 2 analyzes the ventilator data according to the patient digital model to obtain an analysis result;

[0058] A parameter adjustment module 3, the parameter adjustment module 3 is connected to the breathing condition analysis module 2 and the ventilator 2A, and the parameter adjustment module 3 adjusts the control parameters of the ventilator according to the analysis result.

[0059] Specifically, in view of the problem of poor control effect of the ventilator in the prior art, a patient modeling module 1 is set in this embodiment, and a patient digital model is constructed through digital twin technology. Then, the breathing condition analysis module 2 analyzes the ventilator data according to the patient digital model to obtain a more accurate analysis result, so that the subsequent adjustment process of the parameter adjustment module 3 is more in line with the actual situation of the user.

[0060] In one embodiment, as Figure 2 shown, the patient modeling module 1 includes:

[0061] A data acquisition module 11, the data acquisition module 11 extracts sensor data from external sensors;

[0062] The data processing module 12 is connected to the data acquisition module 11. The data processing module preprocesses the sensor data to obtain preprocessed data.

[0063] The modeling module 13 is connected to the data processing module 12. The patient modeling module 13 generates a patient digital model according to the preprocessed data.

[0064] Specifically, aiming at the problem of poor control effect of ventilators in the prior art, in this embodiment, a data acquisition module 11 is set up. The data acquisition module 11 can extract sensor data from the external sensor 1A. This part of the external sensors mainly includes various physiological sensors, such as thermometers, ambient temperature sensors, weighing scales, cameras with pose estimation functions, etc. It can also access external data sources through a collector, such as a medical database, so as to obtain the basic information of the patient, including gender, age, existing diseases, relevant dimensions of the torso, etc.

[0065] Since the above data has multiple sources and different data formats, it needs to be preprocessed by the data processing module 12 after collection, including unifying the format, removing redundant values, filling null values, etc., so as to form preprocessed data in a unified format and establish a table.

[0066] Finally, the modeling module 13 processes the preprocessed data using digital twin technology to establish a patient digital model.

[0067] In one embodiment, as Figure 3 shown, the patient modeling module 13 includes:

[0068] The virtual model retrieval module 131 determines the corresponding model type according to the basic information of the patient in the preprocessed data. The virtual model retrieval module 131 determines the corresponding basic model according to the model type.

[0069] The model adjustment module 132 is connected to the virtual model retrieval module 131. The model adjustment module 132 adjusts the basic model according to the preprocessed data to generate a patient digital model.

[0070] Specifically, to achieve a better digital model generation effect, in this embodiment, first, corresponding basic models are established for different types of patients, mainly in terms of different genders, ages, and weights. This part of the basic models mainly focuses on modeling the torso part and respiratory tract of the patient and endowing them with corresponding physical properties to simulate the lung breathing process under different torso volumes; for other parts, mainly the oxygen consumption of each tissue under different weights and genders is calculated. By constructing this type of basic model, the breathing process can be better predicted, and the oxygen demand of patients under different conditions can be simulated.

[0071] Based on this, the virtual model retrieval module 131 determines the corresponding model type according to the basic patient information in the preprocessed data, and then can retrieve the basic model of the corresponding type through the model type.

[0072] Since there are certain differences between the basic model and the actual patient information, the model adjustment module 132 adjusts the basic model according to the preprocessed data, including changing the body shape-related data and simulating the breathing condition again to generate a patient digital model.

[0073] In one embodiment, as Figure 5 shown, the breathing condition analysis module 2 includes:

[0074] A ventilator interaction module 21, which is connected to the ventilator 2A and collects ventilator data;

[0075] A mapping table adjustment module 22, which is connected to the patient modeling module 1. The mapping table adjustment module 22 adjusts the breathing condition scoring table according to the patient digital model to obtain an adjusted table;

[0076] A condition comparison module 23, which is respectively connected to the ventilator interaction module 21 and the mapping table adjustment module 22. The condition comparison module 23 compares the ventilator data with the adjusted table to obtain an analysis result.

[0077] Specifically, to achieve a better analysis process, in this embodiment, after the patient digital model is established, the mapping table adjustment module 22 adjusts the breathing condition scoring table according to the patient digital model to obtain an adjusted table. The breathing condition scoring table is a scoring table compiled according to existing relevant medical diagnostic indicators, which records the reasonable value ranges corresponding to traditional breathing indicators. By simulating the patient digital model, the table can be effectively adjusted.

[0078] Then, after the ventilator interaction module 21 collects the ventilator data from the ventilator 2A, the condition comparison module 23 compares the ventilator data with the adjusted table to obtain an analysis result.

[0079] In one embodiment, a plurality of ventilator sensors 2A1 are respectively installed on the ventilator 2A. The ventilator sensors 2A1 include a flow sensor, a pressure sensor, and an oxygen sensor;

[0080] The flow sensor is installed in the ventilator pipeline of the ventilator;

[0081] The pressure sensors are respectively installed at the outlet and the return port of the ventilator pipeline of the ventilator;

[0082] The oxygen sensor is installed in the oxygen mask of the ventilator.

[0083] As Figure 4 shown, the ventilator interaction module 21 includes:

[0084] A data acquisition module 211, the data acquisition module 211 is respectively connected to the ventilator sensors 2A1, and the data acquisition module 211 acquires raw data from the ventilator sensors through corresponding data protocols;

[0085] A data format unification module 212, the data format unification module 212 is connected to the data acquisition module 211, and the data format unification module 212 converts the raw data according to the data protocol to form ventilator data.

[0086] Specifically, to achieve better data acquisition effects, in this embodiment, first, the data acquisition module 211 is respectively connected to a flow sensor, a pressure sensor, and an oxygen sensor arranged at specific positions, and then, raw data is acquired from the ventilator sensors through corresponding data protocols, and then the data format unification module 212 converts the raw data according to the data protocol, including format conversion and timing alignment, to form unified ventilator data. This ventilator data is timing data, and information such as gas flow, air supply pressure, return pressure, and oxygen content in the mask at each time point can be obtained by reading this timing data.

[0087] In one embodiment, as Figure 5 shown, the mapping table adjustment module 22 includes:

[0088] A model comparison module 221, the model comparison module 221 obtains a patient digital model and compares the patient digital model with a reference model to form model difference data;

[0089] A respiration prediction module 222, the respiration prediction module 222 is connected to the model comparison module 221, and the respiration prediction module 222 predicts respiration prediction data according to the model difference data;

[0090] A threshold calibration module 223, the threshold calibration module 223 is connected to the respiration prediction module 222, and the threshold calibration module 223 determines a respiration state threshold according to the respiration prediction data and adjusts the respiration condition scoring table to obtain an adjusted table.

[0091] Specifically, to achieve a better table adjustment effect, in this embodiment, first, a patient digital model is compared with a reference model to form model difference data. This part mainly extracts the simulation differences between the two models, including the influence of different thoracic volumes on exhalation volume, the influence of different body weights and blood pressures on oxygen consumption, etc. Then, the respiration prediction module 222 re-predicts according to the model difference data to obtain respiration prediction data, that is, the possible change range of various data during the actual respiration process of the patient. On this basis, the threshold calibration module 223 determines the respiration state threshold according to the respiration prediction data and adjusts the respiration condition score table to obtain an adjusted table.

[0092] Based on the above settings, the respiration condition score table includes two sets of data for each item, that is, the reference numerical range and the percentage-based deviation values at both ends. Then, the respiration state threshold mainly adjusts the reference numerical range and evaluates various conditions through the percentage-based deviation values at both ends.

[0093] In one embodiment, as Figure 6 shown, the condition comparison module 23 includes:

[0094] A preliminary comparison module 231 that compares each item in the ventilator data with the adjusted table respectively to determine the preliminary comparison result;

[0095] A cause prediction module 232 that is connected to the preliminary comparison module 231. The cause prediction module 232 inputs the decision tree model according to the preliminary comparison module 231 to predict and obtain the analysis result.

[0096] Specifically, to achieve a better comparison effect, in this embodiment, after obtaining the ventilator data, first, the preliminary comparison module 231 compares each item in the ventilator data with the adjusted table respectively to determine the relevant indicators whose values deviate from the normal range and outputs them as the preliminary comparison result.

[0097] Since multiple indicators may be abnormal when a patient has difficulty breathing, in this embodiment, a decision tree model is also pre-trained to jointly analyze multiple indicators to determine the actual breathing problem.

[0098] In one embodiment, as Figure 7 shown, the parameter adjustment module 3 includes:

[0099] A parameter optimization module 31 that inputs the analysis result into the parameter optimization model to obtain the model output parameters;

[0100] The model output parameters include control parameters and sensor expected parameters;

[0101] The closed-loop adjustment module 32 is connected to the parameter optimization module 31. The closed-loop adjustment module 32 controls the ventilator according to the control parameters and collects ventilator data.

[0102] The closed-loop adjustment module 32 also adjusts the control parameters according to the ventilator data until the ventilator data meets the sensor expected parameters.

[0103] Specifically, to achieve a better parameter optimization effect, a parameter optimization model constructed based on AI technology is introduced in this embodiment. After determining the breathing problem as the analysis result, the analysis result is input into the parameter optimization model to obtain the model output parameters. The model output parameters include two parts: control parameters and sensor expected parameters. Subsequently, the closed-loop adjustment module 32 first controls the ventilator according to the control parameters and collects the adjusted ventilator data, and gradually adjusts the control parameters according to the real-time ventilator data until the ventilator data meets the sensor expected parameters.

[0104] The above are only the preferred embodiments of the present invention, and do not limit the implementation manners and protection scope of the present invention. For those skilled in the art, it should be able to realize that all the equivalent replacements and obvious changes made by using the description and illustrations of the present invention should be included in the protection scope of the present invention.

Claims

1. A real-time ventilator parameter optimization system based on digital twin technology, comprising: A patient modeling module, wherein the patient modeling module is connected to an external sensor and acquires sensor data of the external sensor, and the patient modeling module establishes a patient digital model according to the sensor data; A respiratory status analysis module, the respiratory status analysis module is connected to the ventilator and collects ventilator data, the respiratory status analysis module is connected to the patient modeling module, and the respiratory status analysis module analyzes the ventilator data according to the patient digital model to obtain an analysis result; A parameter adjustment module is connected to the respiratory condition analysis module and the ventilator, and the parameter adjustment module adjusts the control parameters of the ventilator according to the analysis result.

2. The real-time ventilator parameter optimization system according to claim 1, characterized in that: The patient modeling module includes: a data acquisition module, wherein the data acquisition module extracts the sensor data from the external sensor; A data processing module, the data processing module is connected to the data acquisition module, and the data processing module pre-processes the sensor data to obtain pre-processed data; A modeling module, wherein the modeling module is connected to the data processing module, and the modeling module generates the patient digital model according to the preprocessed data.

3. The real-time ventilator parameter optimization system according to claim 2, characterized in that: The patient modeling module includes: A virtual model retrieval module, wherein the virtual model retrieval module determines a corresponding model type according to the basic information of the patient in the preprocessed data, and the virtual model retrieval module determines a corresponding basic model according to the model type; A model adjustment module, wherein the model adjustment module is connected to the virtual model retrieval module, and the model adjustment module adjusts the basic model according to the preprocessing data to generate the patient digital model.

4. The real-time ventilator parameter optimization system according to claim 1, characterized in that: The respiratory status analysis module comprises: A ventilator interaction module, the ventilator interaction module is connected to the ventilator and collects the ventilator data; A mapping table adjustment module, the mapping table adjustment module is connected to the patient modeling module, and the mapping table adjustment module adjusts the respiratory status score table according to the patient digital model to obtain an adjustment table; A status comparison module is connected to the ventilator interaction module and the mapping table adjustment module respectively, and the status comparison module compares the ventilator data with the adjustment table to obtain the analysis result.

5. The real-time ventilator parameter optimization system according to claim 4, characterized in that: A plurality of ventilator sensors are respectively installed on the ventilator, and the ventilator sensors include a flow sensor, a pressure sensor and an oxygen sensor; The flow sensor is installed in the ventilator pipeline of the ventilator; The pressure sensors are respectively installed in the ventilator pipeline outlet and the ventilator pipeline return port of the ventilator; The oxygen sensor is installed in the oxygen mask of the ventilator.

6. The real-time ventilator parameter optimization system according to claim 5, characterized in that: The ventilator interaction module comprises: A data acquisition module, wherein the data acquisition modules are respectively connected to the ventilator sensors, and the data acquisition modules collect raw data from the ventilator sensors through corresponding data protocols; A data format unification module is connected to the data acquisition module, and the data format unification module converts the raw data according to the data protocol to form the ventilator data.

7. The real-time ventilator parameter optimization system according to claim 4, characterized in that: The mapping table adjustment module includes: A model comparison module, wherein the model comparison module obtains the patient digital model and compares the patient digital model with a reference model to form model difference data; A breathing prediction module, the breathing prediction module is connected to the model comparison module, and the breathing prediction module predicts breathing prediction data according to the model difference data; A threshold calibration module is connected to the breathing prediction module, and the threshold calibration module determines the breathing state threshold according to the breathing prediction data, and adjusts the breathing state scoring table to obtain the adjustment table.

8. The real-time ventilator parameter optimization system according to claim 4, characterized in that: The status comparison module includes: A preliminary comparison module, which compares each item in the ventilator data with the adjustment table to determine a preliminary comparison result; A cause prediction module, wherein the cause prediction module is connected to the preliminary comparison module, and the cause prediction module is input into a decision tree model according to the preliminary comparison module to predict the analysis result.

9. The real-time ventilator parameter optimization system according to claim 1, characterized in that: The parameter adjustment module comprises: A parameter optimization module, wherein the parameter optimization module inputs the analysis results into a parameter optimization model to obtain model output parameters; The model output parameters include control parameters and sensor expected parameters; A closed-loop regulation module, wherein the closed-loop regulation module is connected to the parameter optimization module, and the closed-loop regulation module controls the ventilator according to the control parameters and collects the ventilator data; The closed-loop regulation module further adjusts the control parameters according to the ventilator data until the ventilator data meets the sensor expected parameters.

Images