Identification and Localization of Endotracheal Tube on Chest Radiographs Using a Cascaded Convolutional Neural Network Approach

Abstract

Rapid and accurate assessment of endotracheal tube (ETT) location is essential in the intensive care unit (ICU) setting, where timely identification of a mispositioned support device may prevent significant patient morbidity and mortality. This study proposes a series of deep learning-based algorithms which together iteratively identify and localize the position of an ETT relative to the carina on chest radiographs. Using the open-source MIMIC Chest X-Ray (MIMIC-CXR) dataset, a total of 16,000 patients were identified (8000 patients with an ETT and 8000 patients without an ETT). Three different convolutional neural network (CNN) algorithms were created. First, a regression loss function CNN was trained to estimate the coordinate location of the carina, which was then used to crop the original radiograph to the distal trachea and proximal bronchi. Second, a classifier CNN was trained using the cropped inputs to determine the presence or absence of an ETT. Finally, for radiographs containing an ETT, a third regression CNN was trained to both refine the coordinate location of the carina and identify the location of the distal ETT tip. Model accuracy was assessed by comparing the absolute distance of prediction and ground-truth coordinates as well as CNN predictions relative to measurements documented in original radiologic reports. Upon five-fold cross validation, binary classification for the presence or absence of ETT demonstrated an accuracy, sensitivity, specificity, PPV, NPV, and AUC of 97.14%, 97.37%, 96.89%, 97.12%, 97.15%, and 99.58% respectively. CNN predicted coordinate location of the carina, and distal ETT tip was estimated within a median error of 0.46 cm and 0.60 cm from ground-truth annotations respectively. Overall final CNN assessment of distance between the carina and distal ETT tip was predicted within a median error of 0.60 cm from manual ground-truth annotations, and a median error of 0.66 cm from measurements documented in the original radiology reports. A serial cascaded CNN approach demonstrates high accuracy for both identification and localization of ETT tip and carina on chest radiographs. High performance of the proposed multi-step strategy is in part related to iterative refinement of coordinate localization as well as explicit image cropping which focuses algorithm attention to key anatomic regions of interest.

Automated Endotracheal Tube Placement Check Using Semantically Embedded Deep Neural Networks

“…Deep learning (DL) has been widely used in image processing since the successful application of CNNs to image classification and object recognition (12). Various DL approaches have been applied to many different aspects of CXR, ranging from image quality control ( 13) and detection of the orientation of CXR images (14,15) to more complicated applications, such as anatomic segmentation (16,17), disease classification (18)(19)(20)(21)(22), and ETT detection (23)(24)(25)(26)(27). Several reports have been published for automated detection of the ETT on CXR.…”

“…However, the sensitivity was only 66.5% and specificity 99.2% when assessing ETT-carina distance 7 cm. Using the open-source MIMIC Chest X-Ray dataset and 3 CNN algorithms, Kara et al (25) reported the accuracy, sensitivity, specificity, PPV, NPV, and AUC of 97.14%, 97.37%, 96.89%, 97.12%, 97.15%, and 99.58% respectively, upon a five-fold cross-validation to classify the presence or absence of the ETT, and the model-predicted locations of the carina and the distal tip of ETT were estimated having a median error of 0.46 cm and 0.60 cm from manual groundtruth annotations, respectively. Harris et al (26) trained a bounding box-based CNN for localization of the ET tube and the carina, with a mean difference of 0.7 cm between the model-predicted ETT-carina distance and the ground-truth measurement agreed by two experienced radiologists in 200 CXR images.…”

OOOE: Only-One-Object-Exists Assumption to Find Very Small Objects in Chest Radiographs

“…Additionally, we also provide results on a large scale proprietary dataset of over 100K chest X-ray images. Our method inspired by the OOOE assumption outperforms two commonly used baselines: (1) a simple segmentation model [23,26] and (2) a regression based detection approach [15]. We additionally demonstrate that our approach leads to a model that generalizes better across datasets and makes better use of global context information.…”

“…Regression based approaches such as [15] are trained by directly minimizing the mean-square-error (MSE) between the predicted location g(X) = ŷreg ∈ R 2 and the ground truth location y as depicted in (a) of Figure 1.…”

“…Their method addresses data imbalance and improves performance of an image level classification, but does not provide a solution for the small object detection problem. Kara et al [15] proposed a regression based cascade method to localize the tip of ETT and the carina. In comparison, we provide a classification based solution to the detection problem which is often reported to outperform regression based methods for various detection tasks in natural images [12,19,20,24,27].…”