Tracheal intubation status detection method, computer device and storage medium

Abstract

The present application relates to the field of endotracheal intubation status detection, and specifically to an endotracheal intubation status detection method, a computer device, and a storage medium. A specific implementation of the method includes: in response to receiving an endotracheal intubation status detection instruction for an endotracheal device, controlling a monitoring sensor to monitor the effusion status information inside the oral support device in real time; controlling an associated saliva suction machine to extract effusion; in response to determining that the effusion extraction failed, determining the pressure data of the saliva suction machine's suction tube during effusion extraction; performing abnormal identification on the monitoring video, and in response to determining that the abnormal identification result meets the first abnormal condition, controlling the medical staff prompt device to emit an alarm prompt sound; identifying the position information of the trachea on the slide device in the monitoring video; and adjusting the position of the trachea on the slide device according to the position information and the body posture image. This implementation can timely change the position of the trachea and extract the effusion in the oral cavity, ensuring smooth breathing for the patient.

Description

Technical Field

[0001] The present application relates to the field of endotracheal intubation status detection, and specifically to an endotracheal intubation status detection method, a computer device, and a storage medium. Background Art

[0002] Comatose intubated patients cannot exhale on their own, and long-term bed rest makes nursing care more difficult. They need to turn over frequently to prevent bedsores. The mouth with the tube inserted cannot be closed, resulting in saliva accumulation in the mouth, which requires medical staff to check and remove it in real time. The long-term fixed position of the tube can also cause pressure sores on the patient's face. Turning over sometimes causes the position of the tube to change and often requires changing the position. If the position is not changed in time, it is easy to cause the patient to have difficulty breathing. Summary of the invention

[0003] The content of this application is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this application is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0004] Some embodiments of the present application propose a tracheal intubation status detection method, a computer device, and a computer-readable storage medium to solve the technical problems mentioned in the above background technology section.

[0005] In a first aspect, some embodiments of the present application provide a method for detecting the state of an endotracheal tube, which is applied to an endotracheal device, wherein the endotracheal device comprises: an endotrachea, a slide rail device, an oral support device, a saliva suction machine, a monitoring sensor and a monitoring device, the method comprising: in response to receiving an instruction for detecting the state of an endotracheal tube for the endotracheal device, controlling the monitoring sensor to monitor in real time the effusion state information inside the oral support device, wherein the endotracheal device is in an operating state, in which the slide rail device and the oral support device are placed in the mouth of a target user, a standard interface for a saliva suction machine is left on the outside of the oral support device, the standard interface for saliva suction is connected to the saliva suction machine, and a plurality of pipe through holes are provided inside the oral support device; in response to determining that the effusion state information indicates the presence of effusion, controlling the associated saliva suction machine machine to extract the effusion; in response to determining that the effusion extraction has failed, determining the pressure data of the suction tube of the saliva suction machine when the effusion is extracted; in response to determining that the pressure data is in a preset pressure range, controlling the monitoring device to collect a monitoring video of the oral area of ​​the target user; performing abnormality identification on the monitoring video to obtain an abnormality identification result, and in response to determining that the abnormality identification result meets a first abnormality condition, controlling the associated medical staff prompting device to emit an alarm prompt sound; in response to determining that the abnormality identification result indicates that there is no abnormality, identifying the position information of the trachea on the slide device in the monitoring video, and controlling the monitoring device to collect the current body posture image of the target user; adjusting the position of the trachea on the slide device according to the position information and the body posture image.

[0006] In a second aspect, the present application also provides a computer device, comprising a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, the method described in any implementation of the first aspect is implemented.

[0007] In a third aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, wherein when the computer program is executed by a processor, the method described in any implementation of the first aspect is implemented.

[0008] The above-mentioned embodiments of the present application have the following beneficial effects: the tracheal position can be changed in time, and the effusion in the oral cavity can be extracted, so as to ensure the patient's smooth breathing. First, in response to receiving the tracheal intubation state detection instruction for the tracheal device, the monitoring sensor is controlled to monitor the effusion state information inside the above-mentioned oral support device in real time. Among them, the tracheal device is in an operating state. In the operating state, the above-mentioned slide rail device and the above-mentioned oral support device are placed in the mouth of the target user, and the above-mentioned oral support device has a standard interface for a saliva suction machine on the outside, and the saliva suction standard interface is connected to the saliva suction machine, and the inside of the oral support device is provided with a plurality of pipeline through holes. Thus, it can be determined whether there is effusion inside the patient's oral cavity. Secondly, in response to determining that the above-mentioned effusion state information indicates the presence of effusion, the associated saliva suction machine is controlled to extract the effusion. Thus, the effusion inside the oral cavity can be extracted in time to avoid affecting the patient's breathing. Next, in response to determining that the effusion extraction fails, the pressure data of the suction tube of the above-mentioned saliva suction machine during the effusion extraction is determined. In response to determining that the above-mentioned pressure data is in a preset pressure interval, the above-mentioned monitoring device is controlled to collect the monitoring video of the above-mentioned target user's oral area. Thus, after the extraction of the effusion fails, it is possible to detect whether there is a problem with the straw. Then, the above-mentioned monitoring video is subjected to abnormal identification to obtain an abnormal identification result, and in response to determining that the above-mentioned abnormal identification result satisfies the first abnormal condition, the associated medical staff prompt device is controlled to emit an alarm prompt sound. Thus, when it is detected that there is no abnormality in the internal pressure of the straw, it is possible to detect whether the straw is installed properly through the video image. Afterwards, in response to determining that the above-mentioned abnormal identification result indicates that there is no abnormality, the position information of the trachea on the above-mentioned slide device in the above-mentioned monitoring video is identified, and the above-mentioned monitoring device is controlled to collect the current body posture image of the above-mentioned target user. Finally, according to the above-mentioned position information and the above-mentioned body posture image, the position of the trachea on the above-mentioned slide device is adjusted. Thus, the position of the trachea can be adjusted to ensure smooth breathing. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The above and other features, advantages and aspects of the embodiments of the present application will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar figure labels represent the same or similar elements. It should be understood that the drawings are schematic and that the components and elements are not necessarily drawn to scale.

[0010] Figure 1 is a flow chart of some embodiments of the endotracheal intubation status detection method according to the present application;

[0011] Figure 2 It is a schematic diagram of the structure of a computer device suitable for implementing some embodiments of the present application. DETAILED DESCRIPTION

[0012] Embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are only for exemplary purposes and are not intended to limit the scope of protection of the present application.

[0013] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other.

[0014] It should be noted that the concepts such as "first" and "second" mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0015] It should be noted that the modifications of "one" and "plurality" mentioned in the present application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0016] The names of the messages or information exchanged between multiple devices in the embodiments of the present application are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0017] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments.

[0018] Figure 1 The flowchart of some embodiments of the tracheal intubation state detection method according to the present application is shown. The process 100 of some embodiments of the tracheal intubation state detection method according to the present application is shown. The tracheal intubation state detection method is applied to a tracheal device, and the tracheal device includes: a trachea, a slide rail device, a mouth support device, a saliva suction machine, a monitoring sensor and a monitoring device, and includes the following steps:

[0019] Step 101, in response to receiving a tracheal intubation status detection instruction for a tracheal device, controlling a monitoring sensor to monitor the fluid accumulation status information inside the above-mentioned oral support device in real time.

[0020] In some embodiments, the execution subject (e.g., a computer device) of the tracheal intubation state detection method can control the monitoring sensor to monitor the effusion state information inside the above-mentioned oral support device in real time in response to receiving the tracheal intubation state detection instruction for the tracheal device. Wherein, the tracheal device is in an operating state, and in the operating state, the above-mentioned slide rail device and the above-mentioned oral support device are placed in the mouth of the target user, and the above-mentioned oral support device has a standard interface of a saliva suction machine on the outside, and the standard interface of the saliva suction machine is connected to the saliva suction machine, and the inside of the oral support device is provided with a plurality of pipeline through holes. The trachea can be a respiratory trachea that assists the target user in breathing. The oral support device can be a device that supports the target user to open his mouth. For example, the oral support device can be an oral opener, wherein the trachea passes through the pipeline through hole in the oral support device. The suction tube of the saliva suction machine penetrates from another pipeline through hole inside the oral support device and is placed in the oral cavity of the target user. The monitoring sensor can include an effusion sensor and a humidity sensor. The effusion sensor can detect whether there is effusion in the oral cavity of the target user. The humidity sensor can detect whether the oral cavity of the target user is dry. The monitoring device may be a camera device for photographing the user's face and body posture. For example, the monitoring device may include two cameras, one for photographing the user's face and the other for photographing the user's whole body. The slide rail device may include a trachea connector and a slide rail, wherein one end of the trachea connector is connected to the slide rail and can slide, and the other end of the trachea connector is used to fix the trachea. That is, the position of the trachea can be adjusted by sliding the trachea connector on the slide rail device. The effusion status information may indicate whether there is effusion in the oral cavity of the target user. The slide rail device may be fixed at a position below the nose and above the lips. The standard interface of the saliva suction machine is connected to the saliva suction machine through a straw. The target user refers to the patient.

[0021] Step 102: in response to determining that the effusion state information indicates the presence of effusion, controlling an associated saliva suction machine to extract the effusion.

[0022] In some embodiments, the execution subject may control the associated saliva suction machine to extract the effusion in response to determining that the effusion state information indicates the presence of effusion. For example, the saliva suction machine may be started to operate to suck the effusion in the oral cavity.

[0023] Step 103, in response to determining that the extraction of the effusion has failed, determining the pressure data of the suction tube of the saliva suction machine when the effusion is extracted.

[0024] In some embodiments, the execution subject may determine the pressure data of the suction tube of the saliva suction machine during the extraction of the effusion in response to determining that the extraction of the effusion has failed. It should be noted that when the suction tube of the saliva suction machine is not properly inserted or the suction tube is pressed, the extraction of the effusion may fail. The pressure data may represent the suction pressure value inside the suction tube when the saliva suction machine is working. For example, the pressure data of the suction tube during the extraction of the effusion may be measured by a pressure sensor built into the saliva suction machine.

[0025] Step 104 , in response to determining that the pressure data is within a preset pressure range, controlling the monitoring device to collect monitoring video of the target user's mouth area.

[0026] In some embodiments, the execution subject may control the monitoring device to collect monitoring video of the mouth area of ​​the target user in response to determining that the pressure data is in a preset pressure interval. The preset pressure interval may be a preset suction pressure interval range inside the suction tube when the saliva suction machine is running. That is, the monitoring device may be controlled to collect video of the mouth area of ​​the target user in real time.

[0027] Step 105: perform abnormality identification on the monitoring video to obtain an abnormality identification result, and in response to determining that the abnormality identification result meets the first abnormality condition, control the associated medical staff prompting device to emit an alarm prompt sound.

[0028] In some embodiments, the execution subject may perform abnormal identification on the monitoring video to obtain an abnormal identification result, and in response to determining that the abnormal identification result satisfies the first abnormal condition, control the associated medical staff prompt device to emit an alarm prompt sound. Abnormal identification may include: identification of whether the trachea is folded, identification of whether the straw is folded, identification of whether the straw is fully inserted into the mouth, and identification of whether the trachea is fully inserted into the mouth. For example, the first abnormal condition may be: the abnormal identification result indicates that the trachea is folded, or the abnormal identification result indicates that the straw is folded, or the abnormal identification result indicates that the straw is not fully inserted into the mouth, or the abnormal identification result indicates that the trachea is not fully inserted into the mouth. The medical staff prompt device may be a voice playback device for prompting medical staff. For example, the medical staff prompt device may be a speaker. The alarm prompt sound may be a voice prompting the medical staff that the tracheal device is not installed properly.

[0029] In an actual application scenario, the above execution subject can control the associated medical staff prompt device to emit an alarm prompt sound through the following steps:

[0030] The first step is to perform frame processing on the above surveillance video to obtain a surveillance frame image sequence. The surveillance video can be split into frames of images to obtain a surveillance frame image sequence.

[0031] The second step is to determine the pixel change value between every two frames of the monitoring frame image sequence to obtain a pixel change value sequence. For example, the pixel change value between every two frames of the image can be calculated using an image subtraction formula.

[0032] In the third step, for each pixel change value in the above pixel change value sequence, the following processing steps are performed:

[0033] 1. Determine whether the above pixel change value is greater than or equal to a preset change value.

[0034] 2. In response to determining that the pixel change value is less than or equal to a preset change value, determine the second monitoring frame image corresponding to the pixel change value as the first candidate monitoring frame image.

[0035] 3. In response to determining that the pixel change value is greater than the preset change value, determine the second monitoring frame image corresponding to the pixel change value as the second candidate monitoring frame image.

[0036] The fourth step is to select at least one first alternative monitoring frame image that meets the first target condition from the above-mentioned first alternative monitoring frame images as the third alternative monitoring frame image group. The first target condition can be: the last first alternative monitoring frame image of each first alternative monitoring frame image with a continuous frame number, or a separate first alternative monitoring frame image that is not a continuous frame number. For example, the frame numbers corresponding to each first alternative monitoring frame image can be: 2, 4, 5, 6, 7, 9, 11, 13; then 4, 5, 6, 7 are continuous frame numbers, and the 7th frame image is retained; and images with frame numbers 2, 9, 11, and 13 are retained.

[0037] Step 5: remove the third candidate monitoring frame images that meet the second target condition from the third candidate monitoring frame image group to obtain a fourth candidate monitoring frame image group, wherein the second target condition is: the image frame number is the same as the frame number of any second candidate monitoring frame image in each second candidate monitoring frame image.

[0038] In the sixth step, the fourth candidate monitoring frame image group and the above-mentioned second candidate monitoring frame images are combined into a monitoring frame image group to be detected.

[0039] In the seventh step, the above-mentioned monitoring frame image group to be detected is input into the pre-trained abnormal target recognition model to obtain an abnormal recognition result. Among them, the above-mentioned abnormal recognition result includes: an image abnormality recognition sub-result group, and one image abnormality recognition sub-result corresponds to one monitoring frame image to be detected. The abnormal target recognition model can be a pre-trained target detection model that takes the monitoring frame image to be detected as input and the image abnormality recognition sub-result as output. For example, the abnormal target recognition model can be a YOLOv4 model, a YOLOv5 model, an SSD (Single Shot MultiBox Detector) model, and a DETR (Detection Transformer) model. The YOLOv4 model improves the performance of target detection by improving the network structure, loss function and data enhancement technology. YOLOv4 is characterized by its high precision and good positioning ability, especially in small target detection and dense target detection. As the latest version of the YOLO series, YOLOv5 introduces the ideas of EfficientDet and PANet, realizes a lightweight network structure and multi-level feature fusion, and improves the efficiency and accuracy of target detection. The SSD model can quickly detect objects of different scales by predicting multiple bounding boxes and category probabilities on feature maps of different scales.

[0040] The abnormal target recognition model can be trained by the following steps:

[0041] The first step is to obtain a monitoring frame image training sample set, wherein the monitoring frame image training samples in the monitoring frame image training sample set include: sample monitoring frame images and sample abnormal target labels.

[0042] The second step is to determine the initial abnormal target recognition model. The above-mentioned initial abnormal target recognition model includes: an initial aggregation network, an initial perception network, an initial enhancement network, and an initial recognition convolutional network. For example, the initial aggregation network can be: a TCA (Temporal Context Aggregation) model. The initial perception network can be a neural network model that takes the initial monitoring frame image aggregation information as input and the initial monitoring frame image perception information as output. For example, the initial perception model can be: an MLP (Multilayer Perceptron) model. The initial enhancement network can be a neural network model that takes the initial monitoring frame image perception information as input and the initial monitoring frame image enhancement information as output. For example, the initial enhancement model can be: a PEL (Presupposition Evidence Logic, enhanced prompt) model. The initial recognition convolutional network can be a neural network model that takes the initial user fall behavior image enhancement information as input and the initial abnormal target recognition result as output. For example, the initial recognition convolutional model can be: a CC (causal convolution) model.

[0043] The third step is to select a target monitoring frame image training sample from the monitoring frame image training sample set. A monitoring frame image training sample can be randomly selected from the monitoring frame image training sample set as the target monitoring frame image training sample.

[0044] The fourth step is to input the sample monitoring frame images included in the target monitoring frame image training samples into the initial aggregation network to obtain initial monitoring frame image aggregation information.

[0045] The fifth step is to input the initial monitoring frame image aggregation information into the initial perception network to obtain the initial monitoring frame image perception information.

[0046] In the sixth step, the initial monitoring frame image perception information is input into the initial enhancement network to obtain the initial monitoring frame image enhancement information.

[0047] In the seventh step, the initial monitoring frame image enhancement information is input into the initial recognition convolutional network to obtain the initial abnormal target recognition result.

[0048] In the eighth step, based on a preset loss function, the recognition loss value between the initial abnormal target recognition result and the corresponding sample abnormal target label is determined. For example, the preset loss function may be a mean square error loss function (MSE), a hinge loss function (SVM), or a cross entropy loss function (CrossEntropy).

[0049] In a ninth step, in response to determining that the recognition loss value is less than or equal to the preset recognition loss value, the initial abnormal target recognition model is determined as the trained abnormal target recognition model.

[0050] Thus, through the initial aggregation network, the initial monitoring frame image aggregation information taking into account the time and contextual relationship can be more accurately identified. Then, through the initial perception network, the feature information in the initial monitoring frame image aggregation information can be more accurately identified. Next, through the initial enhancement network, more accurate initial monitoring frame image enhancement information taking into account the user's oral area can be obtained. After that, through the initial recognition convolution network, a more accurate initial abnormal target recognition result taking into account the time sequence can be obtained. Therefore, a more accurate abnormal target recognition result can be obtained by training an abnormal target recognition model including an initial aggregation network, an initial perception network, an initial enhancement network, and an initial recognition convolution network. Thus, through a more accurate abnormal target recognition result, it is possible to identify whether the trachea is folded, whether the straw is folded, whether the straw is fully inserted into the oral cavity, and whether the trachea is fully inserted into the oral cavity.

[0051] Step 106, in response to determining that the abnormal identification result indicates no abnormality, identifying the position information of the trachea on the slide rail device in the monitoring video, and controlling the monitoring device to collect the current body posture image of the target user.

[0052] In some embodiments, the execution subject may, in response to determining that the abnormal identification result indicates no abnormality, identify the position information of the trachea on the slide rail device in the monitoring video, and control the monitoring device to collect the current body posture image of the target user. The position information may indicate whether the trachea is at the left corner of the target user's mouth / right corner of the mouth or in the middle, and whether the extension of the trachea is biased to the left or to the right.

[0053] In an actual application scenario, the execution subject can identify the position information of the air pipe on the slide rail device in the monitoring video through the following steps:

[0054] The first step is to perform frame processing on the above monitoring video to obtain a monitoring frame image sequence.

[0055] In the second step, for each monitoring frame image in the above monitoring frame image sequence, the following processing steps are performed:

[0056] 1. Perform multi-scale feature extraction on the monitoring frame image through a multi-scale feature extraction network to generate facial image features. The multi-scale feature extraction network adopts a feature pyramid network FPN (Feature Pyramid Network, FPN).

[0057] 2. Determine the trachea position information based on the facial image features and the first positioning model. The trachea position information includes: mouth confidence, mouth features and trachea key point position group. The first positioning model uses MobileNet-SSD as the basic network architecture to determine the mouth confidence and mouth features. Based on the mouth features, the execution subject combines the SIFT (Scale Invariant Feature Transform) algorithm to determine the trachea key point position group.

[0058] 3. In response to the above-mentioned mouth confidence being greater than the preset confidence, determine the candidate positioning point position information according to the above-mentioned mouth features and the second positioning model. Wherein, the above-mentioned candidate stimulation point position information includes: the first candidate positioning point position and the second candidate positioning point position, and the above-mentioned multi-scale feature extraction network, the above-mentioned first positioning model and the above-mentioned second positioning model are included in the positioning point positioning model. The above-mentioned multi-scale feature extraction network, the above-mentioned first positioning model and the above-mentioned second positioning model are included in the positioning point positioning model. In practice, the second positioning model and the above-mentioned first positioning model can adopt the same network structure. Specifically, the first positioning model is obtained by training in a supervised manner using image samples marked with the mouth position. The second positioning model is obtained by training in a supervised manner using image samples marked with the trachea positioning point position.

[0059] 4. Determine the first slope based on the above trachea key point position group. The first slope can be determined based on the above trachea key point position group by a two-point equation.

[0060] 5. Determine the second slope based on the first candidate positioning point position and the second candidate positioning point position. The second slope can be determined by a two-point equation based on the first candidate positioning point position and the second positioning stimulation point position.

[0061] 6. Determine the target angle according to the first slope and the second slope. Wherein, tan(β)=abs((k1-k2) / (1+k1×k2)). Wherein, β is the target angle. k1 is the first slope. k2 is the second slope.

[0062] 7. In response to the target angle being smaller than a preset angle, the candidate positioning point position information is determined as candidate position information.

[0063] The third step is to determine the middle position of each candidate position information as the position information.

[0064] Step 107, adjusting the position of the trachea on the slide rail device according to the position information and the body posture image.

[0065] In some embodiments, the execution subject may adjust the position of the trachea on the slide rail device according to the position information and the body posture image. For example, if the body posture image indicates that the user is lying on the left side, the corresponding trachea position interval information should be to the right of the middle corner of the mouth. That is, if the position information indicates that the trachea is at the left corner of the mouth, the slide rail on the slide rail device is controlled to slide the trachea to the right corner of the mouth. For another example, if the body posture image indicates that the user is lying on the right side, the corresponding trachea position interval information should be to the left of the middle corner of the mouth. That is, if the position information indicates that the trachea is at the right corner of the mouth, the slide rail on the slide rail device is controlled to slide the trachea to the left corner of the mouth.

[0066] In an actual application scenario, the execution subject can adjust the position of the air pipe on the slide rail device through the following steps:

[0067] In the first step, multi-scale image feature extraction is performed on the above-mentioned body posture image to generate a body posture image feature set. The first convolution layer, the second convolution layer and the third convolution layer can be used to extract image features from the above-mentioned body posture image to obtain a body posture image feature set. Among them, the convolution kernel of the first convolution layer is 2×2. The convolution kernel of the second convolution layer is 3×3. The convolution kernel of the third convolution layer is 1×1.

[0068] The second step is to generate fused image features based on the above body posture image feature set. The body posture image features can be fused through the FilterConcatenation layer to obtain fused image features.

[0069] The third step is to generate body posture information for the body posture image based on the fused image features. In practice, the fused image features can be input into a pre-trained body posture recognition model to obtain body posture information. The body posture recognition model includes: an aggregation network, a perception network, an enhancement network, and a body posture recognition convolutional network.

[0070] The fourth step is to determine the trachea position interval information corresponding to the above body posture information. For example, if the body posture image indicates that the user is lying on the left side, the corresponding trachea position interval information should be to the right of the middle of the mouth corner. For another example, if the body posture image indicates that the user is lying on the right side, the corresponding trachea position interval information should be to the left of the middle of the mouth corner.

[0071] In a fifth step, in response to determining that the position information is not within the trachea position interval information, the slide rail device is controlled to slide the position of the trachea to within the trachea position interval information.

[0072] Further, in response to determining that the above-mentioned fluid accumulation state information indicates that the humidity of the oral cavity is less than a preset threshold, the associated nebulizer is controlled to perform nebulization.

[0073] In some embodiments, the execution subject may control the associated sprayer to spray in response to determining that the fluid accumulation state information indicates that the oral humidity is less than a preset threshold. The sprayer may be a sprayer placed above the corner of the target user's mouth, and spray when the user's mouth is dry.

[0074] The present application also provides a computer device 200. Figure 2 As shown, the computer device 200 includes: a bus 201, a processor 202, a memory 203 and a communication interface 204. The processor 202, the memory 203 and the communication interface 204 communicate with each other through the bus 201. The computer device 200 can be a server or a terminal device. It should be understood that the present application does not limit the number of processors and memories in the computer device 200.

[0075] The bus 201 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 2 The bus 201 is represented by only one line, but it does not mean that there is only one bus or one type of bus. The bus 201 may include a path for transmitting information between various components of the computer device 200 (eg, the memory 203, the processor 202, and the communication interface 204).

[0076] The processor 202 may include any one or more of a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP).

[0077] The memory 203 may include a volatile memory, such as a random access memory (RAM). The memory 203 may also include a non-volatile memory, such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD), or a solid state drive (SSD).

[0078] The memory 203 stores executable program codes, and the processor 202 executes the executable program codes to respectively implement the functions of the aforementioned acquisition module, sampling module, determination module, and mixing module, thereby implementing the aforementioned method for detecting the state of the tracheal intubation. That is, the memory 202 stores instructions for executing the aforementioned method for detecting the state of the tracheal intubation.

[0079] The communication interface 204 uses a transceiver module such as, but not limited to, a network interface card or a transceiver to implement communication between the computer device 200 and other devices or communication networks.

[0080] An embodiment of the present application also provides a chip, which includes a processor and a data interface. The processor reads instructions stored in a memory through the data interface to execute the above-mentioned method for detecting the state of tracheal intubation.

[0081] The embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that can be stored by a computing device or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk). The computer-readable storage medium includes instructions that instruct the computing device to execute the above-mentioned method for detecting the state of the endotracheal tube.

[0082] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0083] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the protection scope of the technical solutions of the embodiments of the present application.

Claims

1. A tracheal device, using a tracheal intubation status detection method, the tracheal device comprising: Trachea, slide rail device, mouth support device, saliva suction machine, monitoring sensor and monitoring device, the method comprises: In response to receiving a tracheal intubation status detection instruction for the tracheal device, the monitoring sensor is controlled to monitor the effusion status information inside the mouth support device in real time, wherein the tracheal device is in an operating state, in which the slide rail device is fixed at a position below the nose and above the lips of the target user, the mouth support device is placed in the mouth of the target user, a standard interface for a saliva suction machine is reserved on the outside of the mouth support device, the standard interface for saliva suction is connected to the saliva suction machine, and a plurality of pipeline through holes are arranged inside the mouth support device; In response to determining that the effusion status information indicates the presence of effusion, controlling an associated saliva suction machine to extract the effusion; In response to determining that the extraction of the effusion has failed, determining pressure data of the suction tube of the saliva suction machine during the extraction of the effusion; In response to determining that the pressure data is within a preset pressure interval, controlling the monitoring device to collect monitoring video of the target user's mouth area; Performing abnormality identification on the surveillance video to obtain an abnormality identification result, and in response to determining that the abnormality identification result meets a first abnormality condition, controlling an associated medical staff prompting device to emit an alarm prompt sound; In response to determining that the abnormal identification result indicates no abnormality, identifying the position information of the trachea on the slide rail device in the monitoring video, and controlling the monitoring device to collect a current body posture image of the target user; The position of the trachea on the slide rail device is adjusted according to the position information and the body posture image.

2. The tracheal device according to claim 1, characterized in that: The method further comprises: In response to determining that the fluid accumulation state information indicates that the oral humidity is less than a preset threshold, the associated nebulizer is controlled to perform nebulization.

3. The tracheal device according to claim 1, characterized in that: The performing abnormality identification on the surveillance video to obtain an abnormality identification result includes: Performing frame processing on the monitoring video to obtain a monitoring frame image sequence; Determine the pixel change value between every two frames of the monitoring frame image sequence to obtain a pixel change value sequence; For each pixel change value in the pixel change value sequence, the following processing steps are performed: Determine whether the pixel change value is greater than or equal to a preset change value; In response to determining that the pixel change value is less than or equal to a preset change value, determining a second monitoring frame image corresponding to the pixel change value as a first candidate monitoring frame image; In response to determining that the pixel change value is greater than the preset change value, determining a second monitoring frame image corresponding to the pixel change value as a second candidate monitoring frame image; Selecting at least one first candidate monitoring frame image that meets the first target condition from each first candidate monitoring frame image as a third candidate monitoring frame image group; Removing the third candidate monitoring frame images that meet the second target condition from the third candidate monitoring frame image group, to obtain a fourth candidate monitoring frame image group, wherein the second target condition is: the image frame number is the same as the frame number of any second candidate monitoring frame image in each of the second candidate monitoring frame images; Combining the fourth candidate monitoring frame image group and each of the second candidate monitoring frame images into a monitoring frame image group to be detected; The monitoring frame image group to be detected is input into a pre-trained abnormal target recognition model to obtain an abnormal recognition result, wherein the abnormal recognition result includes: an image abnormality recognition sub-result group, and one image abnormality recognition sub-result corresponds to one monitoring frame image to be detected.

4. The tracheal device according to claim 1, characterized in that: The step of adjusting the position of the trachea on the slide rail device according to the position information and the body posture image includes: Performing multi-scale image feature extraction on the body posture image to generate a body posture image feature set; generating fused image features according to the body posture image feature set; generating body posture information for the body posture image according to the fused image features; Determining trachea position interval information corresponding to the body posture information; In response to determining that the position information is not within the air pipe position interval information, the slide rail device is controlled to slide the position of the air pipe to within the air pipe position interval information.

5. The tracheal device according to claim 4, characterized in that: Generating body posture information for the body posture image according to the fused image features includes: The fused image features are input into a pre-trained body posture recognition model to obtain body posture information, wherein the body posture recognition model includes: an aggregation network, a perception network, an enhancement network, and a body posture recognition convolutional network.

6. A computer device, wherein: The computer device includes a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, the steps of the tracheal intubation status detection method applied to the tracheal equipment as described in any one of claims 1-5 are implemented.

7. A computer-readable storage medium, wherein: The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the steps of the tracheal intubation status detection method applied to the tracheal device as described in any one of claims 1-5 are implemented.

Images