Warning device for misplacement of endotracheal tube or tracheostomy tube

Abstract

To provide a warning device for an abnormal position of an endotracheal tube.SOLUTION: A warning device for an abnormal position of an endotracheal tube that monitors accuracy of an installation position of an endotracheal tube includes: a monitoring module 1 for collecting chest X-ray image materials of a patient; an object detection module 2 for detecting an object based on a deep learning model; a position evaluation module 3 for interpreting the position appropriateness of the endotracheal tube based on a result of the object detection module; and a display module 4 for displaying an evaluation result on the position appropriateness of the endotracheal tube and providing a necessary warning as a clinical reference. Thereby, the position of the endotracheal tube is evaluated by an artificial intelligence calculation model using a chest X-ray image captured for the patient. A main objective is to enable clinical medical staff to quickly and accurately determine the appropriateness of the position of the endotracheal tube, reduce the rate of unplanned slipping of the endotracheal tube and the incidence of single-lung ventilation, and improve the medical quality, thereby improving the safety and vital sign stability of the patient.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a warning device for abnormal position of an endotracheal tube. In particular, it relates to medical images, and further, it can assist clinical medical staff to quickly and accurately confirm whether the position of the endotracheal tube is appropriate, and by reducing the unplanned slippage rate of the endotracheal tube and the incidence of one-lung ventilation, improving medical quality, it is related to enhancing patient safety and vital sign stability.

Background Art

[0002] The unplanned slippage rate of the endotracheal tube (ETT) and the incidence of one-lung ventilation are not high, but when they occur, the patient's vital signs become unstable, the length of hospital stay and medical expenses increase, and in the worst case, injuries such as death occur. Therefore, the accuracy of the endotracheal tube position becomes more important clinically. Also, in the medical environment, for patients in intensive care units (ICUs), on average, a chest X-ray is taken every two to five days, and now it is necessary for doctors to manually interpret it. Therefore, when clinical doctors examine chest X-rays under a huge workload, they are always attracted by the main lesions such as pneumonia, and can only examine the position of the endotracheal tube one by one.

[0003] In the intensive care unit, occasionally, unplanned slippage rates of the endotracheal tube and one-lung ventilation events occur, causing fatal injuries to patients. In theory, the position of the endotracheal tube should be carefully examined for each chest X-ray. However, because doctors are very busy, under a huge workload, they usually pay attention to other obvious lesions, so the dislocation of the endotracheal tube may be ignored. Therefore, there is a need for a device that can detect and judge the position of the endotracheal tube and issue a warning to assist doctors' judgment.

Disclosure of the Invention

Problems to be Solved by the Invention

[0004] The main objective of the present invention is to provide an artificial intelligence (AI) computational model that can solve the above-mentioned problems of the conventional approach and automatically identify the position of the endotracheal tube in chest X-ray images. This enables clinical medical staff to quickly and accurately confirm whether the endotracheal tube is in the correct position, thereby reducing the incidence of unplanned endotracheal tube slippage and unilateral lung ventilation, improving medical quality, and enhancing patient safety and vital sign stability.

[0005] Another object of the present invention is to provide an artificial intelligence computational model that can automatically interpret the appropriate position of an endotracheal tube in a chest X-ray image and issue an appropriate warning, and to provide an endotracheal tube misposition warning device that exhibits excellent performance in both the development and testing phases, and improves the quality of patient care and safety in clinical practice.

[0006] Another object of the present invention is to take one chest X-ray image every two to five days on average for patients in the intensive care unit to evaluate the appropriateness of the endotracheal tube position, train an artificial intelligence computational model using multiple chest X-ray images, and then use the artificial intelligence computational model to 1) determine the position of the tracheal carina in the chest X-ray image to determine whether an endotracheal tube or tracheostomy tube is being used and to determine its position, and 2) calculate the appropriateness of the endotracheal tube position and warn if the endotracheal tube position is abnormal.

[0007] Another object of the present invention relates to the field of image processing technology, and in particular, to a method for distance scaling of image analysis and marking. In image processing applications, it is common to calculate and measure the distance between objects in an image to perform analysis and marking. However, since the size and aspect ratio of different images may differ, directly displaying the distance using pixel values ​​may be inaccurate. Therefore, there is a need for a method that can perform accurate image analysis and marking by converting the pixel values ​​of an image into real-world distance values.

[0008] The method includes: step 1 defining the original size of the training dataset image, which includes the height and width of the image; step 2 receiving an input image awaiting processing; step 3 calculating a scaling r, which is the ratio of the size of the input image to the size of the training dataset image, and selecting a relatively small value as the scaling; step 4 converting the pixel values ​​of the input image to actual distance values ​​corresponding to the training dataset image based on the calculated scaling r; and step 5 using the converted actual distance values ​​for image analysis and marking, for example, calculating the distance between objects and marking them. [Means for solving the problem]

[0009] To achieve the above objectives, the present invention provides an endotracheal tube position abnormality warning device for monitoring the accuracy of the placement of an endotracheal tube, comprising: a monitoring module for collecting chest X-ray image data of a patient; and a device connected to the monitoring module that receives the chest X-ray image data input from the monitoring module and uses an artificial intelligence (AI) computational model to identify the tracheal carina, endotracheal tube (ETT), and tracheostomy tube in the image. The system includes: an object detection module that identifies the presence of tracheal carinae or endotracheal tubes and annotates their location; a location evaluation module connected to the object detection module that, when the object detection module detects the presence of a tracheal carinae or endotracheal tube, automatically measures the distance between the tip of the endotracheal tube and the tracheal carinae based on the annotation location of the judgment result of the artificial intelligence calculation model, generates an evaluation result for the appropriateness of the position of the endotracheal tube based on the calculated distance, and triggers an alarm if the position of the endotracheal tube is not within the correct range in the evaluation result; and a display module connected to the location evaluation module that receives the judgment results of the object detection module and the location evaluation module, issues a warning signal, and displays the annotation result of the object and the evaluation result of the appropriateness of the position of the endotracheal tube.

[0010] According to the above embodiments of the present invention, the object detection module, based on a YOLOv5 deep learning model, inputs chest X-ray image data, and uses multiple chest X-ray image data at preset ratios as a training dataset and a test dataset for deep learning, respectively. The module performs model training on the training dataset and tests on the test dataset, and then terminates the artificial intelligence computation model by verification. The performance of the artificial intelligence computation model is then evaluated using standard function evaluation metrics.

[0011] According to the embodiments described above of the present invention, the standard function evaluation index includes, but is not limited to, precision, recall, mean average precision (mAP@50), and accuracy.

[0012] According to the above embodiments of the present invention, the position evaluation module triggers an alarm if the tip position of the endotracheal tube is lower than the tracheal carina, or if the distance between the tip of the endotracheal tube and the tracheal carina is less than 3 cm or greater than 5 cm.

[0013] According to the above embodiments of the present invention, the chest X-ray image data of the object detection module includes chest X-rays of endotracheal tubes, chest X-rays with tracheostomy tubes, and chest X-rays with or without endotracheal tubes.

[0014] According to the embodiments described above of the present invention, the source of chest X-rays with the endotracheal tube is primarily images of actual patients requiring endotracheal tube treatment, which include cases where the placement of the endotracheal tube is appropriate or inappropriate, and the patient group includes patients with different endotracheal tube experiences and thoracic anatomical structures. [Best Mode for Carrying Out the Invention]

[0015] Referring to Figures 1 to 4F, the following are diagrams: a conceptual diagram of the structure of the endotracheal tube position abnormality warning device according to the present invention; a conceptual diagram of the usage flow according to the present invention; a conceptual diagram of when the artificial intelligence calculation model according to the present invention determines that there is no endotracheal tube; a conceptual diagram of when the artificial intelligence calculation model according to the present invention determines that the position of the endotracheal tube is accurate; a conceptual diagram of when the artificial intelligence calculation model according to the present invention determines that the position of the endotracheal tube is abnormal; a conceptual diagram of when the artificial intelligence calculation model according to the present invention determines that the position of the endotracheal tube is within an abnormal distance range; an image diagram of when the artificial intelligence calculation model according to the present invention takes the shortest distance of the three determined distances; and when the artificial intelligence calculation model according to the present invention detects the position of the tracheal carina. The images show the following: an image when the endotracheal tube is not detected, an image when the artificial intelligence calculation model according to the present invention detects the tracheal carina and tracheostomy tube, an image when the artificial intelligence calculation model according to the present invention detects the tracheal carina and endotracheal tube and determines that the tip position of the endotracheal tube is within the normal distance range, an image when the artificial intelligence calculation model according to the present invention detects the tracheal carina and endotracheal tube and determines that the distance between the tip position of the endotracheal tube and the tracheal carina is less than 3 cm, and an image when the artificial intelligence calculation model according to the present invention detects the tracheal carina and endotracheal tube and determines that the distance between the tip position of the endotracheal tube and the tracheal carina is greater than 5 cm. As shown in the figure, the present invention is an endotracheal tube position abnormality warning device that monitors the accuracy of the installation position of the endotracheal tube, and consists of a monitoring module 1, an object detection module 2, a position evaluation module 3, and a display module 4.

[0016] The above-mentioned monitoring module 1 collects chest X-ray images of the patient.

[0017] The property detection module 2 is connected to the monitoring module 1 and detects properties based on a YOLO V5 deep learning model.

[0018] The position evaluation module 3 is connected to the object detection module 2, and determines the appropriateness of the position of the endotracheal tube based on the results from the object detection module 2.

[0019] The display module 4 is connected to the position evaluation module 3 and displays the evaluation results of the appropriate position of the endotracheal tube, and also issues a warning as a clinical reference as necessary. The above structure constitutes a novel endotracheal tube positional abnormality warning device 100.

[0020] The distance measurement process described above is performed by the positional distance suitability calculation module, and source code inspection confirms that at least two objects are detected, and that the first object is the tracheal carina and the second object is the endotracheal tube, or vice versa. If a tracheal carina or endotracheal tube is detected, the source code reads the corresponding coordinate information from the detected object. The source code then calculates the distance between the endotracheal tube and the tracheal carina. The calculation process involves 1) calculating the straight-line distance between the two objects using the Euclidean distance formula, and 2) scaling the predicted image and training dataset image using the calculated distance value, while simultaneously converting the pixel value to the actual distance value. For the tip position of the endotracheal tube, the leftmost and rightmost coordinate points are selected from the bottom edge of the annotation bounding box for the identified endotracheal tube tip. Then, a function is used to calculate the midpoint between the two points. For example, points A and C in Figure 4A are the rightmost and leftmost points at the bottom of the bounding box, and point B is the midpoint at the bottom of the bounding box, i.e., the midpoint between points A and C. The distance value from the tracheal carina (for example, point D in Figure 4A) is calculated using points A, B, and C, and the shortest distance is taken as the result. The relevant coordinate points and distance marks, along with the distance value and warnings, are then displayed on the image.

[0021] Figures 4B to 4F show examples of results determined by the artificial intelligence computation model according to the present invention. In Figure 4B, the model detected the location of the tracheal carina but did not detect the endotracheal tube. Figure 4C shows that the model detected the tracheal carina and tracheostomy tube. Figure 4D shows that the model detected the tracheal carina and endotracheal tube and that the tip position of the endotracheal tube is within the normal distance range. Figure 4E shows that the model detected the tracheal carina and endotracheal tube and that the distance between the tip position of the endotracheal tube and the tracheal carina is less than 3 cm. Figure 4F shows that the model detected the tracheal carina and endotracheal tube and that the distance between the tip position of the endotracheal tube and the tracheal carina is greater than 5 cm.

[0022] According to the present invention, the object detection module 2 uses an artificial intelligence (AI) computational model to identify and annotate the tracheal carina, endotracheal tube (ETT), and tracheostomy tubes in the chest X-ray image data input from the monitoring module 1. When the object detection module 2 detects the presence of a tracheal carina or endotracheal tube, the position evaluation module 3 automatically measures the distance between the tip of the endotracheal tube and the tracheal carina based on the annotation position determined by the AI ​​computational model. Based on the obtained distance value, it generates an evaluation result of the appropriateness of the endotracheal tube's position. If the evaluation result indicates that the endotracheal tube is not located within the correct range, an alarm is triggered. Finally, the display module 4 receives the determination results from the object detection module 2 and the position evaluation module 3, issues a warning signal, and displays the object annotation result and the evaluation result of the appropriateness of the endotracheal tube's position. Therefore, this device 100 can be applied to a wide range of medical-related industries, and in order to reduce the workload of clinical physicians, the present invention uses artificial intelligence technology to detect and determine the position of the endotracheal tube, and provides this as important reference information for the appropriateness of the tracheal position to the physician's decision-making.

[0023] According to the present invention, the object detection module 2 and position evaluation module 3 for evaluating the appropriateness of the position of the endotracheal tube utilize chest X-ray images taken of the patient and evaluate the position of the endotracheal tube using an advanced artificial intelligence computation model. The main purpose is to enable clinical medical staff to quickly and accurately determine whether the position of the endotracheal tube is appropriate, thereby reducing the incidence of unplanned endotracheal tube slippage and unilateral lung ventilation, and thereby improving the quality of medical care, patient safety, and vital sign stability.

[0024] According to the present invention, the object detection module 2 is trained with a YOLO V5 model using 2278 image samples, and then tested and verified with 253 images. By predicting the position of the tracheal carina in the X-ray image, it checks for the presence of an endotracheal tube or tracheostomy tube. If an endotracheal tube is detected, the position evaluation module 3 further calculates the distance between the tip of the endotracheal tube and the tracheal carina, and at the same time, the problem of the endotracheal tube being set too deep can be eliminated. Standard functional evaluation metrics include, but are not limited to, precision, recall, mean average precision (mAP@50), and accuracy, ensuring high accuracy of the model.

[0025] The distance scaling formula described above calculates the ratio between the pixel values ​​of the input image and the predicted image relative to the training dataset images. It then converts both of these ratios into distance values ​​for practical applications, enabling accurate image analysis and marking. Because this method is simple and easy to implement, it can be widely applied in image processing applications, improving the accuracy and efficiency of image analysis.

[0026] According to the present invention, the object detection module 2 and the position evaluation module 3 do not require any special devices. Moreover, due to the compatibility with existing chest X-ray image acquisition devices and medical systems, a simple integration solution can be obtained. Clinical staff can quickly obtain the evaluation result of the position of the endotracheal tube by simply inputting the pixel material of the patient's X-ray into the information system.

[0027] According to a better specific embodiment of the present invention, the chest X-ray pixel material of the object detection module 2 includes chest X-rays of endotracheal tubes, chest X-rays with tracheostomy tubes, and chest X-rays without endotracheal tubes or tracheostomy tubes.

[0028] According to a better specific embodiment of the present invention, the source of the chest X-ray with an endotracheal tube is mainly the image of an actual patient who requires treatment of the endotracheal tube, including cases where the position of the endotracheal tube is appropriate or inappropriate, and the patient group includes patients with different endotracheal tube experiences and chest anatomies.

[0029] According to a better specific embodiment of the present invention, the position evaluation module 3 triggers an alarm when the tip position of the endotracheal tube is lower than the tracheal carina, or when the distance between the tip of the endotracheal tube and the tracheal carina is less than 3 cm or greater than 5 cm.

[0030] Hereinafter, the embodiments will illustrate the details and meaning of the present invention by way of examples, but the scope of the claims of the present invention is not limited thereby.

[0031] According to a better embodiment, the source of the test material is the actual chest X-ray image of the patient. It includes the normal situation of the endotracheal tube position and the incorrect position of the inner tube. Moreover, it includes different cases and diversifications of different patients. Since the data is anonymized, the privacy and compliance of the patient are protected. Also, patients with different endotracheal tube experiences and chest anatomies are included.

[0032] The total number of chest X-ray images used in the test is 253.

[0033] The test results are, (1) Sensitivity is 0.963, (2) The specificity is 0.964, (3) Average average precision (mAP@50): 0.966, (4) The accuracy is 0.962.

[0034] In a better example, the flow is as shown in Figure 2.

[0035] Model autopsy (YOLOv5) step s11 involves using a pre-trained YOLOv5 model to detect chest X-ray images and identify the tracheal carina and the locations of the endotracheal tube and tracheostomy tube. As shown in Figure 4B, the location of the tracheal carina was detected, but the image does not show an endotracheal tube.

[0036] Step s12, in the model, automatically identifies the tip position of the endotracheal tube and automatically measures the distance between the tip of the endotracheal tube and the tracheal carina.

[0037] Step s13, which determines whether the endotracheal tube is positioned correctly, involves determining whether the endotracheal tube is positioned appropriately based on the tip position of the endotracheal tube and the measured distance between the tip of the endotracheal tube and the tracheal carina, as shown in Figure 3B.

[0038] Step s14, which provides an early warning, warns that the endotracheal tube may be in the wrong position if the tip of the endotracheal tube is lower than the tracheal carina, or if the measured distance is outside the range of 3-5 cm, as shown in Figure 3C.

[0039] Figures 4B-4F show how an artificial intelligence computation model identifies the presence and location of the tracheal carina, endotracheal tube, and tracheostomy tube in the image, and measures the distance between the tip of the endotracheal tube and the tracheal carina. Simultaneously, based on past clinical standards and literature, a standard rule for the endotracheal tube trigger alarm is proposed. According to this rule, as shown in Figure 3B, a green light illuminates when the distance is 3-5 cm; as shown in Figure 3D, a yellow light illuminates when the distance is 2.5-3 cm or 5-5.5 cm; and as shown in Figure 3C, a red light illuminates when the distance is less than 2.5 cm or greater than 5.5 cm. In another simplified example, as shown in Figure 3C, a red light illuminates when the distance is less than 3 cm or greater than 5 cm. Furthermore, if only the tracheal carina is detected and the endotracheal tube is not detected, it is displayed as shown in Figure 3A. Furthermore, according to a better implementation, if the initial alarm is triggered, a flashing image or sound may be used to enhance the attention-grabbing effect, and the flashing or sound may be stopped once the clinical staff understands.

[0040] As can be seen from the above, in critically ill patients using mechanical ventilation, unplanned events such as extubation or unilateral ventilation occasionally occur, which can be fatal to the patient. Theoretically, it is necessary to carefully examine the position of the endotracheal tube in each chest X-ray image. However, endotracheal tube displacement may be overlooked as the physician's attention is drawn to other obvious lesions. Therefore, this invention develops an artificial intelligence computational model that can automatically identify the position of the endotracheal tube on chest X-ray images. The materials and methods used are to establish training and test datasets by randomly selecting anonymized chest X-ray images of critically ill patients, and to establish an artificial intelligence computational model using the Python YOLOv5 model, thereby predicting the position of the tracheal carina in the X-ray image and confirming the presence of an endotracheal tube or tracheostomy tube. If an endotracheal tube is detected, the distance between the tracheal carina and the tip of the endotracheal tube is further calculated, and an alarm is triggered when the tip of the endotracheal tube is located below the tracheal carina, or when the distance is less than 3 cm or greater than 5 cm. According to the present invention, based on experimental results, the accuracy, recall, mean-average accuracy, and precision of the artificial intelligence computational model, using 2278 chest X-ray images as a training dataset and 253 chest X-ray images as a test dataset, are 0.963, 0.964, 0.966, and 0.962, respectively. As external verification, the present invention deploys the artificial intelligence computational model in clinical practice. The alarm system triggered by the artificial intelligence computational model shows a decrease in the median duration (interquartile range) of inappropriate endotracheal tube placement from 3.00 (1.25-4.00) to 2.00 (1.00-3.00). Therefore, the present invention proposes an artificial intelligence computational model that can automatically interpret the appropriateness of endotracheal tube placement in chest X-ray images and issue appropriate warnings. Furthermore, the artificial intelligence computational model achieved excellent performance in both the development and testing phases, and its application to clinical practice improves patient care quality and safety.

[0041] The application scope of this invention is medical imaging, and its main purpose is to enable clinical medical staff to quickly and accurately determine the appropriate position of the endotracheal tube, thereby reducing the incidence of unplanned endotracheal tube slippage and unilateral lung ventilation, and improving the stability of the patient's vital signs.

[0042] 1. According to the above-described warning device for abnormal position of an endotracheal tube, the artificial intelligence calculation module for evaluating the appropriateness of the position of the endotracheal tube is required to have the following uses and effects.

[0043] 1. Automated evaluation is performed on the position of the endotracheal tube in the patient's chest X-ray image. 2. To help clinical medical staff quickly and accurately determine whether endotracheal tube placement is appropriate. 3. Reduce the incidence of unplanned endotracheal tube slippage and unilateral lung ventilation, thereby lowering the risk of related complications and improving medical quality, patient safety, and vital sign stability.

[0044] 2. The indications for use of the endotracheal tube abnormality warning device according to the present invention are as follows:

[0045] 1. The indications for using the above artificial intelligence computing module include, but are not limited to, its use in determining whether the position of the endotracheal tube is appropriate when applied to chest X-ray images taken from a patient. 2. This applies to, but is not limited to, all patients requiring endotracheal tube treatment, including critically ill patients and surgical patients.

[0046] For patients in the intensive care unit, on average, a chest X-ray is taken every two to five days, and then it awaits manual interpretation by a physician. However, when clinicians examine chest X-rays under an enormous workload, their attention is sometimes drawn to the main lesions such as pneumonia, making it impossible to confirm the position of each endotracheal tube. Therefore, this invention is an endotracheal tube position abnormality warning device that selects / detects the position of the patient's endotracheal tube and provides an evaluation result immediately. It is an auxiliary tool that can primarily provide endotracheal tube alerts to clinical medical staff, confirm the proper position of the endotracheal tube, improve the treatment outcome for the patient, and reduce associated risks.

[0047] As described above, the technical features of the present invention are as follows. 1. Train using the YOLO V5 model. We will test and verify with 253 images, using 2278 sample images for training. 2. Predict the X-ray image. a. While confirming the location of the tracheal carina, check for the presence of an endotracheal tube or tracheostomy tube. b. If an endotracheal tube is present, the distance between the tracheal carina and the tip of the endotracheal tube is further calculated to determine the appropriateness of the endotracheal tube's position. 3. Proof of concept.

[0048] As described above, the present invention is a warning device for abnormal tracheal tube positioning that effectively overcomes the shortcomings of conventional devices. For patients in the intensive care unit, one chest X-ray image is taken every two to five days on average to evaluate the appropriateness of the tracheal tube position. Furthermore, an artificial intelligence (AI) computational model is trained on multiple chest X-ray image samples. The AI ​​computational model then determines 1) the position of the tracheal carina in the chest X-ray image, whether an endotracheal tube (ETT) or tracheostomy tube is being used, and its position, and 2) calculates the appropriateness of the tracheal tube position and issues a warning. Therefore, the present invention is more progressive and practical, and can reliably satisfy the needs of users. Accordingly, we propose patent claims in accordance with the law.

[0049] The foregoing is merely a better embodiment of the present invention, and the scope of the invention is not limited thereto, and various equivalent changes and modifications are included within the claims of the present invention. [Brief explanation of the drawing]

[0050] [Figure 1] This is a conceptual diagram of the structure of a warning device for abnormal positioning of an endotracheal tube according to the present invention. [Figure 2] This is a conceptual diagram of the usage flow according to the present invention. [Figure 3A] This is a conceptual diagram of the artificial intelligence computational model according to the present invention when it determines that there is no endotracheal tube. [Figure 3B] This is a conceptual diagram of how the artificial intelligence computational model according to the present invention determines that the position of the endotracheal tube is accurate. [Figure 3C] This is a conceptual diagram of when the artificial intelligence computational model according to the present invention determines that the position of the endotracheal tube is abnormal. [Figure 3D] This is a conceptual diagram showing the state in which the artificial intelligence computational model according to the present invention determines that the position of the endotracheal tube is nearly abnormal. [Figure 4A]This is an image diagram showing the process of selecting the shortest distance from three determined distances using the artificial intelligence computation model according to the present invention. [Figure 4B] This image shows the results of using the artificial intelligence computational model according to the present invention, where the position of the tracheal carina was detected, but the endotracheal tube was not detected. [Figure 4C] This image shows the results of detecting a tracheal carina and tracheostomy tube using the artificial intelligence computational model according to the present invention. [Figure 4D] This image shows the result of using the artificial intelligence computational model according to the present invention to detect the tracheal carina and endotracheal tube, and to determine that the tip position of the endotracheal tube is within the normal distance range. [Figure 4E] This image shows the result of using the artificial intelligence computational model according to the present invention to detect the tracheal carina and endotracheal tube, and to determine that the distance between the tip of the endotracheal tube and the tracheal carina is less than 3 cm. [Figure 4F] This image shows the result of using the artificial intelligence computational model according to the present invention to detect the tracheal carina and endotracheal tube, and to determine that the distance between the tip of the endotracheal tube and the tracheal carina is greater than 5 cm. [Explanation of Symbols]

[0051] 1. Monitoring Module Property Detection Module 100 Endotracheal tube misplacement warning device 2. Position evaluation module 3 Display Module 4 Steps 11-14

Claims

1. A warning device for abnormal placement of an endotracheal tube or tracheostomy tube, which monitors the accuracy of the placement of the endotracheal tube, A monitoring module that collects chest X-ray images of patients, A device detection module connected to the aforementioned monitoring module receives image data of the chest X-ray input from the monitoring module, and uses an artificial intelligence (AI) computational model to identify and annotate the location of the tracheal carina, endotracheal tube (ETT), and tracheostomy tubes in the image. A position evaluation module is connected to the aforementioned object detection module, and when the object detection module detects the presence of a tracheal carina and an endotracheal tube or tracheostomy tube, it automatically calculates the actual distance between the tip of the endotracheal tube or tracheostomy tube and the tracheal carina using the annotation position which is the judgment result of the artificial intelligence calculation model, generates an evaluation result of the appropriateness of the position of the endotracheal tube or tracheostomy tube based on the obtained actual distance, and triggers an alarm when the position of the endotracheal tube or tracheostomy tube is not within the correct range. The system includes a display module connected to the location evaluation module, which receives the judgment results from the object detection module and the location evaluation module, issues a warning signal, and displays the object annotation results and the evaluation results of the appropriateness of the position of the endotracheal tube or tracheostomy tube, The position evaluation module includes a positional distance appropriateness calculation module that determines the actual distance between the endotracheal tube or tracheostomy tube and the tracheal carina. The positional distance suitability calculation module selects the leftmost and rightmost coordinate points from the lower end of the annotation boundary box of the tip of the endotracheal tube or tracheostomy tube identified by the object detection module, calculates the midpoint between the two points using a function, calculates the image distance value between these three points and the tracheal carina using the Euclidean distance formula, selects the shortest calculated distance value, and converts the image pixel values ​​to the corresponding actual distance for this distance value. The position evaluation module triggers an alarm when the tip of the endotracheal tube or tracheostomy tube is lower than the tracheal carina, or when the actual distance between the tip of the endotracheal tube or tracheostomy tube and the tracheal carina is less than 3 cm or greater than 5 cm. A warning device for abnormal positioning of an endotracheal tube or tracheostomy tube, characterized by the above features.

2. In the warning device for abnormal position of an endotracheal tube or tracheostomy tube according to Claim 1, The aforementioned property detection module, based on a YOLOv5 deep learning model, inputs chest X-ray image data and, using preset ratios, uses multiple chest X-ray image data sets as training and test datasets for the deep learning model. The model is trained using the training dataset and tested using the test dataset for verification, thereby completing the artificial intelligence computation model. The performance of the artificial intelligence computation model is then evaluated using standard performance evaluation metrics. A warning device for abnormal positioning of an endotracheal tube or tracheostomy tube, characterized by the above features.

3. In the warning device for abnormal position of an endotracheal tube or tracheostomy tube according to Claim 2, The standard performance evaluation metrics include one or more of the following: precision, recall, mean average precision (mAP@50), and accuracy. A warning device for abnormal positioning of an endotracheal tube or tracheostomy tube, characterized by the above features.

4. In the warning device for abnormal position of an endotracheal tube or tracheostomy tube according to Claim 1, The method for triggering the aforementioned alarm is expressed by one or more of the following: illustration, flashing illustration, and audio. A warning device for abnormal positioning of an endotracheal tube or tracheostomy tube, characterized by the above features.

5. In the warning device for abnormal position of an endotracheal tube or tracheostomy tube according to Claim 2, The image data input as training and test datasets for deep learning in order to construct the artificial intelligence computation model of the object detection module is an X-ray image of the chest with an endotracheal tube or tracheostomy tube, and an X-ray image of the chest without an endotracheal tube or tracheostomy tube. A warning device for abnormal positioning of an endotracheal tube or tracheostomy tube, characterized by the above features.

6. In the warning device for abnormal position of an endotracheal tube or tracheostomy tube according to Claim 5, The chest X-ray images of the endotracheal tube or tracheostomy tube are primarily chest X-ray images of actual patients requiring endotracheal tube or tracheostomy tube treatment, and include a determination of whether the position of the endotracheal tube or tracheostomy tube is appropriate. The patient group also includes patients with different endotracheal or tracheostomy tube experiences and thoracic anatomy. A warning device for abnormal positioning of an endotracheal tube or tracheostomy tube, characterized by the above features.

Images