Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Aug 7;12(8):1913.
doi: 10.3390/diagnostics12081913.

Detecting Endotracheal Tube and Carina on Portable Supine Chest Radiographs Using One-Stage Detector with a Coarse-to-Fine Attention

Liang-Kai Mao  1 , Min-Hsin Huang  2 , Chao-Han Lai  2 , Yung-Nien Sun  1 , Chi-Yeh Chen  1
Affiliations

Detecting Endotracheal Tube and Carina on Portable Supine Chest Radiographs Using One-Stage Detector with a Coarse-to-Fine Attention

Liang-Kai Mao et al. Diagnostics (Basel). .

Abstract

In intensive care units (ICUs), after endotracheal intubation, the position of the endotracheal tube (ETT) should be checked to avoid complications. The malposition can be detected by the distance between the ETT tip and the Carina (ETT-Carina distance). However, it struggles with a limited performance for two major problems, i.e., occlusion by external machine, and the posture and machine of taking chest radiographs. While previous studies addressed these problems, they always suffered from the requirements of manual intervention. Therefore, the purpose of this paper is to locate the ETT tip and the Carina more accurately for detecting the malposition without manual intervention. The proposed architecture is composed of FCOS: Fully Convolutional One-Stage Object Detection, an attention mechanism named Coarse-to-Fine Attention (CTFA), and a segmentation branch. Moreover, a post-process algorithm is adopted to select the final location of the ETT tip and the Carina. Three metrics were used to evaluate the performance of the proposed method. With the dataset provided by National Cheng Kung University Hospital, the accuracy of the malposition detected by the proposed method achieves 88.82% and the ETT-Carina distance errors are less than 5.333±6.240 mm.

Keywords: coarse-to-fine attention; deep learning; endotracheal intubation; object detection.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Figures

Figure 1
Figure 1
An overview of the proposed architecture for detecting the malposition. It consists of a ResNet-based backbone, Coarse-to-Fine Attention (CTFA), FPN-based neck, FCOS-based detection head, and segmentation head. The legends below demonstrate the operations above.
Figure 2
Figure 2
An illustration of coarse-to-fine attention (CTFA). CTFA consisted of a global-modelling attention (GA) and a scale attention (SA). GA was aimed at capturing long-range relationships and SA was aimed at reweighting with local relationships.
Figure 3
Figure 3
An illustration of global-modelling attention (GA). GA generated long-range relationships through two branches. The upper branch was aimed at capturing long-range context information and the lower branch was aimed at grabbing local context information. Then, this information is integrated by a series of operations.
Figure 4
Figure 4
An illustration of scale attention (SA). SA addressed the defects of convolutional block attention module (CBAM) by adaptive channel pooling and squeeze-and-excitation (SE) block.
Figure 5
Figure 5
An illustration of post-process.
Figure 6
Figure 6
Ground Truth. (a) Original ground truth. (b) Pre-processed ground truth.
Figure 7
Figure 7
Ensuring at most one ETT tip/Carina left. (a) Without post-process. (b) With post-process.
Figure 8
Figure 8
Refining the feature point of ETT tip/Cairna by the bbox of ETT/Bifurcation. (a) Without post-process. (b) With post-process.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Hunter T.B., Taljanovic M.S., Tsau P.H., Berger W.G., Standen J.R. Medical devices of the chest. Radiographics. 2004;24:1725–1746. doi: 10.1148/rg.246045031. - DOI - PubMed
    1. Varshney M., Sharma K., Kumar R., Varshney P.G. Appropriate depth of placement of oral endotracheal tube and its possible determinants in Indian adult patients. Indian J. Anaesth. 2011;55:488. doi: 10.4103/0019-5049.89880. - DOI - PMC - PubMed
    1. Bentz M.R., Primack S.L. Intensive care unit imaging. Clin. Chest Med. 2015;36:219–234. doi: 10.1016/j.ccm.2015.02.006. - DOI - PubMed
    1. Chen S., Zhang M., Yao L., Xu W. Endotracheal tubes positioning detection in adult portable chest radiography for intensive care unit. Int. J. Comput. Assist. Radiol. Surg. 2016;11:2049–2057. doi: 10.1007/s11548-016-1430-3. - DOI - PubMed
    1. Kao E.F., Jaw T.S., Li C.W., Chou M.C., Liu G.C. Automated detection of endotracheal tubes in paediatric chest radiographs. Comput. Methods Programs Biomed. 2015;118:1–10. doi: 10.1016/j.cmpb.2014.10.009. - DOI - PubMed

LinkOut - more resources