Ventilation trigger identification method and system of anesthesia machine and anesthesia machine

Abstract

The invention relates to the technical field of medical instruments, in particular to a ventilation trigger recognition method and system of an anaesthesia machine and the anaesthesia machine. The method comprises the following steps: step 1, collecting airway pressure waveform data and flow waveform data of the anaesthesia machine; 2, converting the collected airway pressure waveform data and flow waveform data into matrix data; 3, inputting the matrix data into a trained convolutional neural network, wherein the convolutional neural network outputs a probability value based on the input matrix data; 4, comparing the probability value with a threshold value, and judging whether a ventilation triggering condition is met or not; if the triggering condition is met, the anesthesia machine starts ventilation support. The shape of the ventilation curve is analyzed by adopting the convolutional neural network, and the triggering condition can be more accurately judged, so that more accurate and safer anesthesia ventilation control is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of medical devices, and in particular to a ventilation trigger identification method and system for an anesthesia machine and an anesthesia machine. Background Art

[0002] In the ventilation control system of the anesthesia machine, triggered ventilation recognition is a crucial function. This function is particularly important in ventilation modes with trigger functions such as pressure support (PS) and synchronized intermittent mandatory ventilation (SIMV), especially when the user has the ability to breathe spontaneously. When the user initiates an inspiratory effort, the airway pressure will drop, accompanied by changes in the inspiratory flow rate. The anesthesia machine can determine whether the trigger conditions are met by monitoring changes in airway pressure or inspiratory flow. Once the airway pressure drops significantly or the inspiratory flow rate reaches a certain threshold, the device will determine that the user has initiated an inspiratory action, thereby starting ventilation support for the user.

[0003] Traditional anesthesia machines usually follow the following steps to determine whether a trigger has occurred: (1) Determination of the end of the exhalation process: During the exhalation process, the anesthesia machine first needs to determine whether the exhalation has ended. Normally, the device considers the stability of airway pressure and expiratory flow as a sign of the end of exhalation. Only when these two parameters tend to stabilize will the system start the trigger judgment function. Specifically, the end-expiratory pressure is considered to be a sign of the end of exhalation. When the pressure and flow reach a stable state, the exhalation action is considered to be basically completed. (2) Real-time judgment of trigger conditions: After the end of exhalation, the anesthesia machine will determine whether the trigger conditions are met by real-time monitoring of the airway pressure or flow. The specific judgment criteria are as follows:

[0004] Pressure trigger: The device records the end-expiratory pressure value. When the airway pressure is lower than the set end-expiratory pressure value, it is considered that the user has initiated an inhalation effort and the trigger condition is met.

[0005] Flow trigger: When the inhalation flow exceeds the preset threshold, the system determines that the user has inhaled and the trigger condition is met.

[0006] However, in actual use, when the anesthesia machine determines whether the trigger conditions are met, it often triggers falsely or fails to trigger. For example, when the ventilation circuit leaks, the airway pressure may continue to drop after the end of exhalation. Since the traditional system relies on the stability of the airway pressure to determine whether the exhalation is over, the leak may cause the system to misjudge that the exhalation is not over, and thus fail to start the trigger judgment function. Another example is false triggering or failure to trigger due to flow sensor failure or offset. In some cases, the flow sensor may fail or offset, causing it to mistakenly measure the flow when there is no actual flow. This may occur at the end of true exhalation, when the flow sensor may mistakenly show that there is still flow in the air circuit, causing the system to think that the user is still exhaling and unable to start the trigger judgment. On the contrary, if the flow sensor mistakenly detects the inhalation flow, even if the user does not initiate inhalation, the system will mistakenly judge it as an inhalation action, causing a false trigger.

[0007] These problems indicate that the traditional trigger judgment method may not be able to provide accurate and reliable ventilation support in the face of complex or abnormal use environments. Therefore, it is urgent to improve the existing technology to ensure that the user's inspiratory effort can be accurately judged in various situations, so as to achieve safer and more effective anesthesia ventilation control. Summary of the invention

[0008] The purpose of the present invention is to overcome the above-mentioned defects of the prior art, thereby providing a ventilation trigger identification method and system for an anesthesia machine and an anesthesia machine.

[0009] In order to solve the above technical problems, the technical solution of the present invention provides a ventilation trigger identification method for an anesthesia machine, comprising:

[0010] Step 1: Collect airway pressure waveform data and flow waveform data of the anesthesia machine;

[0011] Step 2: converting the collected airway pressure waveform data and flow waveform data into matrix data;

[0012] Step 3: Input the matrix data into a trained convolutional neural network, and the convolutional neural network outputs a probability value based on the input matrix data;

[0013] Step 4: Compare the probability value with the threshold value and determine whether the ventilation trigger condition is met; if the trigger condition is met, the anesthesia machine starts ventilation support.

[0014] As an improvement of the above method, in step 1, airway pressure waveform data and flow waveform data are collected every 100 ms.

[0015] As an improvement to the above method, the matrix data in step 2 is a 20×20 matrix.

[0016] As an improvement of the above method, the trained convolutional neural network in step 3 includes two convolutional layers, two sampling layers and one fully connected layer.

[0017] As an improvement of the above method, the convolution kernel size of the convolution layer is 5×5.

[0018] As an improvement of the above method, the convolutional neural network module is trained by preprocessed ventilation trigger case waveform data, and the ventilation trigger case waveform data includes: volume-controlled pressure trigger waveform data, volume-controlled flow trigger waveform data, pressure-controlled pressure trigger waveform data and pressure-controlled flow trigger waveform data.

[0019] As an improvement of the above method, the preprocessing step of the trigger case waveform data includes: determining the waveform data before ventilation triggering in the trigger case waveform data and retaining it, and setting the waveform data after ventilation triggering to zero.

[0020] To achieve another object of the present invention, the present invention further provides a ventilation trigger recognition system for an anesthesia machine, which is used to execute the ventilation trigger recognition method for an anesthesia machine, comprising:

[0021] A data acquisition module is used to collect airway pressure waveform data and flow waveform data of the anesthesia machine in real time;

[0022] A data conversion module, used for converting the collected airway pressure waveform data and flow waveform data into matrix data;

[0023] Convolutional neural network, which outputs probability values ​​based on the input matrix data;

[0024] The judgment module is used to compare the probability value with the threshold value and judge whether the ventilation trigger condition is met; if the trigger condition is met, the anesthesia machine starts ventilation support.

[0025] As an improvement of the above system, the system also includes: a training module, wherein the training module is used to preprocess the ventilation trigger case waveform data and use the preprocessed ventilation trigger case waveform data for training; the trigger case waveform data includes: volume control pressure trigger waveform data, volume control flow trigger waveform data, pressure control pressure trigger waveform data and pressure control flow trigger waveform data; the preprocessing step of the trigger case waveform data includes: determining the waveform data before ventilation triggering in the trigger case waveform data and retaining it, and setting the waveform data after ventilation triggering to zero.

[0026] To achieve another object of the present invention, the present invention further provides an anesthesia machine, comprising the ventilation trigger identification system of the anesthesia machine mentioned above.

[0027] Compared with the prior art, the advantage of the present invention is that the ventilation trigger recognition method, system and anesthesia machine of the present invention can more accurately judge the trigger conditions by adopting a convolutional neural network to analyze the shape of the ventilation curve instead of simply relying on the absolute value of the sensor, and effectively solves the false triggering or non-triggering caused by problems such as sensor error or circuit leakage in traditional methods, thereby achieving more accurate and safe anesthesia ventilation control. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 A schematic diagram of a ventilation trigger identification method for an anesthesia machine provided in Example 1 of the present invention;

[0029] Figure 2 Schematic diagram of convolutional neural network;

[0030] FIG3( a ) is a schematic diagram of the capacitive pressure triggering waveform data;

[0031] FIG3( b ) is a schematic diagram of the capacity control flow trigger waveform data;

[0032] FIG3( c ) is a schematic diagram of pressure-controlled pressure triggering waveform data;

[0033] FIG3( d ) is a schematic diagram of pressure-controlled flow trigger waveform data;

[0034] Figure 4 Schematic diagram of ventilation trigger judgment steps. DETAILED DESCRIPTION

[0035] The technical solution provided by the present invention is further illustrated below in conjunction with embodiments.

[0036] In order to overcome the problem of false triggering or non-triggering that may be caused by relying on the absolute value of the sensor in the traditional method, the present invention adopts a convolutional neural network (CNN) to determine whether the triggering conditions are met by identifying the shape of the ventilation curve, thereby effectively solving problems such as circuit leakage, pressure sensor offset and flow sensor offset.

[0037] Example 1

[0038] This embodiment provides a ventilation trigger recognition method for an anesthesia machine. Figure 1 As shown, the following steps are included:

[0039] 1. Construction and training of convolutional neural networks

[0040] like Figure 2 As shown in Figure 2, the convolutional neural network (CNN) in this method consists of the following layers:

[0041] Two convolutional layers: used for feature extraction.

[0042] Two sampling layers: used for data compression.

[0043] One fully connected layer: used for the final classification judgment.

[0044] The data input size is set to 20 × 20 and the convolution kernel size is 5 × 5. This configuration can effectively extract and compress data features while ensuring that the data sample size is moderate and the training speed is reasonable.

[0045] 2. Training of Convolutional Neural Networks

[0046] The convolutional neural network is trained using clinical data, and the training process mainly includes the following four types of trigger cases: volume-controlled pressure trigger waveform data as shown in Figure 3(a), volume-controlled flow trigger waveform data as shown in Figure 3(b), pressure-controlled pressure trigger waveform data as shown in Figure 3(c), and pressure-controlled flow trigger waveform data as shown in Figure 3(d).

[0047] During the training process, the convolutional neural network needs to learn the features before the trigger. In order to prevent the waveform after the trigger from interfering with the feature extraction, the ventilation waveform after the trigger is manually removed during training. Specifically, the area of ​​inspiratory effort (i.e. the waveform part before the trigger) is manually determined, and the waveform feature values ​​of the part after the trigger are uniformly set to zero, thereby preventing subsequent waveforms from interfering with the learning process of the device.

[0048] 3. Trigger judgment steps

[0049] Figure 4 The trigger judgment step is shown. In the trigger judgment process, the real-time pressure and flow waveform data are first obtained. Each data sampling point is 100ms, that is, 10 points are collected per second, a total of 20 points are collected, and the data size is 20×20. The matrix data is sufficient to describe the inhalation characteristics before triggering. In this way, the system can complete effective data acquisition, and the memory occupancy of the data matrix is ​​moderate, which will not affect the normal operation of the system. When the system memory space is sufficient, the number of sampling points can be appropriately increased. Then the pressure and flow waveforms are filled with data. The real-time collected pressure and flow data are filled into a 20×20 matrix, and the remaining blank parts are filled with zeros to complete the data extraction of the convolutional neural network. This processing method ensures that the system can effectively extract features from the curve shape and make correct trigger judgments.

[0050] Example 2

[0051] This embodiment provides a ventilation trigger recognition system for an anesthesia machine, which is used to execute the ventilation trigger recognition method for an anesthesia machine of Embodiment 1, including:

[0052] A data acquisition module is used to collect airway pressure waveform data and flow waveform data of the anesthesia machine in real time;

[0053] A data conversion module, used for converting the collected airway pressure waveform data and flow waveform data into matrix data;

[0054] Convolutional neural network, which outputs probability values ​​based on the input matrix data;

[0055] The judgment module is used to compare the probability value with the threshold value and judge whether the ventilation trigger condition is met; if the trigger condition is met, the anesthesia machine starts ventilation support.

[0056] Preferably, the system also includes: a training module, wherein the training module is used to preprocess the ventilation trigger case waveform data and use the preprocessed ventilation trigger case waveform data for training; the trigger case waveform data includes: volume control pressure trigger waveform data, volume control flow trigger waveform data, pressure control pressure trigger waveform data and pressure control flow trigger waveform data; the preprocessing step of the trigger case waveform data includes: determining the waveform data before ventilation triggering in the trigger case waveform data and retaining it, and setting the waveform data after ventilation triggering to zero.

[0057] Example 3

[0058] This embodiment provides an anesthesia machine, including the ventilation trigger recognition system of the anesthesia machine of Embodiment 2.

[0059] From the above specific description of the present invention, it can be seen that the present invention uses a convolutional neural network to analyze the shape of the ventilation curve instead of simply relying on the absolute value of the sensor, so as to more accurately judge the triggering conditions, effectively solving the false triggering or non-triggering caused by problems such as sensor errors or loop leakage in traditional methods, thereby achieving more accurate and safe anesthesia ventilation control.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention is described in detail with reference to the embodiments, it should be understood by those skilled in the art that any modification or equivalent replacement of the technical solutions of the present invention does not depart from the spirit and scope of the technical solutions of the present invention and should be included in the scope of the claims of the present invention.

Claims

1. A ventilation trigger recognition method for an anesthesia machine, comprising: Step 1: Collect airway pressure waveform data and flow waveform data of the anesthesia machine; Step 2: converting the collected airway pressure waveform data and flow waveform data into matrix data; Step 3: Input the matrix data into a trained convolutional neural network, and the convolutional neural network outputs a probability value based on the input matrix data; Step 4: Compare the probability value with the threshold value and determine whether the ventilation triggering condition is met; If the trigger conditions are met, the anesthesia machine starts ventilation support.

2. The ventilation trigger identification method of an anesthesia machine according to claim 1, characterized in that: In step 1, airway pressure waveform data and flow waveform data are collected every 100 ms.

3. The ventilation trigger identification method of an anesthesia machine according to claim 1, characterized in that: The matrix data in step 2 is a 20×20 matrix.

4. The ventilation trigger identification method of an anesthesia machine according to claim 1, characterized in that: The trained convolutional neural network in step 3 includes two convolutional layers, two sampling layers and one fully connected layer.

5. The ventilation trigger identification method of an anesthesia machine according to claim 4, characterized in that: The convolution kernel size of the convolution layer is 5×5.

6. The ventilation trigger identification method of an anesthesia machine according to claim 1, characterized in that: The convolutional neural network module is trained by preprocessed ventilation trigger case waveform data, and the ventilation trigger case waveform data includes: volume control pressure trigger waveform data, volume control flow trigger waveform data, pressure control pressure trigger waveform data and pressure control flow trigger waveform data.

7. The ventilation trigger identification method of an anesthesia machine according to claim 6, characterized in that: The preprocessing step of the trigger case waveform data includes: determining the waveform data before ventilation triggering in the trigger case waveform data and retaining it, and setting the waveform data after ventilation triggering to zero.

8. A ventilation trigger recognition system for an anesthesia machine, used to execute the ventilation trigger recognition method for an anesthesia machine according to any one of claims 1 to 7, comprising: A data acquisition module is used to collect airway pressure waveform data and flow waveform data of the anesthesia machine in real time; A data conversion module, used for converting the collected airway pressure waveform data and flow waveform data into matrix data; Convolutional neural network, which outputs probability values ​​based on the input matrix data; A judgment module, used to compare the probability value with the threshold value and judge whether the ventilation triggering condition is met; If the trigger conditions are met, the anesthesia machine starts ventilation support.

9. The ventilation trigger recognition system for an anesthesia machine according to claim 8, characterized in that: Also includes: A training module, wherein the training module is used to preprocess ventilation trigger case waveform data and use the preprocessed ventilation trigger case waveform data for training; the trigger case waveform data includes: volume control pressure trigger waveform data, volume control flow trigger waveform data, pressure control pressure trigger waveform data and pressure control flow trigger waveform data; the preprocessing step of the trigger case waveform data includes: determining the waveform data before ventilation triggering in the trigger case waveform data and retaining it, and setting the waveform data after ventilation triggering to zero.

10. An anesthesia machine, characterized in that: A ventilation trigger identification system for an anesthesia machine comprising the anesthesia machine described in claim 8 or 9.

Images