Computer-Aided Assessment of Catheters and Tubes on Radiographs: How Good is Artificial Intelligence for Assessment?

@article{Yi2020ComputerAidedAO,
  title={Computer-Aided Assessment of Catheters and Tubes on Radiographs: How Good is Artificial Intelligence for Assessment?},
  author={Xin Yi and Scott J. Adams and Robert D. E. Henderson and Paul S. Babyn},
  journal={Radiology. Artificial intelligence},
  year={2020},
  volume={2 1},
  pages={
          e190082
        }
}
Catheters are the second most common abnormal finding on radiographs. The position of catheters must be assessed on all radiographs because serious complications can arise if catheters are malpositioned. However, due to the large number of radiographs obtained each day, there can be substantial delays between the time a radiograph is obtained and when it is interpreted by a radiologist. Computer-aided approaches hold the potential to assist in prioritizing radiographs with potentially… 

Figures and Tables from this paper

Artificial Intelligence Solutions for Analysis of X-ray Images

Although comprehensive AI solutions across modalities have yet to be developed, institutions can begin to select and integrate focused solutions which increase efficiency, increase quality and patient safety, and add value for their patients.

Classification of Central Venous Catheter Tip Position on Chest X-ray Using Artificial Intelligence

This study aimed to develop an algorithm for the automatic classification of proper depth with the application of automatic segmentation of the trachea and the CVC on chest radiographs using a deep CNN.

Automatic classification of multiple catheters in neonatal radiographs with deep learning

A convolutional neural network trained using a dataset of 777 neonatal chest and abdominal radiographs achieves strong performance in the simultaneous detection of these four catheter types, and may be used as a time-saving mechanism to automate reporting of catheters on radiographs.

Automatic Detection and Classification of Multiple Catheters in Neonatal Radiographs with Deep Learning

A convolutional neural network was trained using a dataset of 777 neonatal chest and abdominal radiographs, with a split of 81%–9%–10% for training-validation-testing, respectively, and achieves strong performance in the simultaneous detection of these four catheter types.

Pediatric chest radiograph interpretation: how far has artificial intelligence come? A systematic literature review

All publicly available pediatric datasets were identified and their potential utility and limitations for pediatric AI studies were determined and the literature was systematically reviewed to assess the current state of AI in pediatric chest radiograph interpretation.

Tubular shape aware data generation for segmentation in medical imaging

This work presents an approach for synthetic generation of the tube-shaped objects, with a generative adversarial network being regularized with a prior-shape constraint, using Frangi-based regularization to draw synthetic tubes in the predefined fake mask regions and uses the adversarial component to preserve the global realistic appearance of the synthesized image.

Development of Automatic Endotracheal Tube and Carina Detection on Portable Supine Chest Radiographs using Artificial Intelligence

A feature extraction method with Mask R-CNN is proposed to overcome the problem of low image quality of portable supine chest radiographs and have a strong correlation between the board-certified intensivists and the result in terms of ETT-carina distance.

Landmark Constellation Models For Central Venous Catheter Malposition Detection

A HRNet-based key point detection approach in combination with a probabilistic constellation model is proposed which enables the exact localization of the CVC tip, but also of relevant anatomical landmarks and provides a likelihood score for tip position which allows us to identify malpositioned CVCs.

Tubular Shape Aware Data Generation for Semantic Segmentation in Medical Imaging

This work presents an approach for synthetic data generation of the tube-shaped objects, with a generative adversarial network being regularized with a prior-shape constraint, and eliminates the need for paired image--mask data and requires only a weakly-labeled dataset to reach the accuracy of the fully-supervised models.

References

SHOWING 1-10 OF 49 REFERENCES

Catheter detection and classification on chest radiographs: an automated prototype computer-aided detection (CAD) system for radiologists

The results from the prototype system show that it is feasible to automatically detect both catheters on chest radiographs, with the potential to significantly speed the delivery of imaging services while maintaining high accuracy.

Automatic Catheter and Tube Detection in Pediatric X-ray Images Using a Scale-Recurrent Network and Synthetic Data

An automatic approach for detection of catheters and tubes on pediatric X-ray images with a UNet-style segmentation network with a recurrent module that can process inputs at multiple scales and iteratively refine the detection result is proposed.

Semi-automated location identification of catheters in digital chest radiographs

This work uses forward looking registration with normalized cross-correlation in order to take advantage of a priori information of the catheter profile to track catheters appearing in digital chest radiographs acquired by portable chest x-rays using a semi-automatic method.

An approach for chest tube detection in chest radiographs

This is the first study conducted to automatically detect artificial objects in the lung region of chest radiographs, and the authors obtained promising results on the detection of artificial objects.

An improved automatic computer aided tube detection and labeling system on chest radiographs

The results from the prototype system show that it is feasible to automatically detect both tubes on chest radiographs, with the potential to significantly speed the delivery of imaging services while maintaining high accuracy.

Deep learning in chest radiography: Detection of findings and presence of change

Deep learning algorithm in its present version is unlikely to replace radiologists due to its limited specificity for categorizing specific findings, however, it can aid in interpretation of CXR findings and their stability over follow up CxR.

Assessment of Critical Feeding Tube Malpositions on Radiographs Using Deep Learning

DCNNs may allow for more rapid identification and communication of critical feeding tube malpositions on radiographs, and deep learning demonstrates promise in differentiating critical vs. non-critical placement.

Automatic Guidewire Tip Segmentation in 2D X-ray Fluoroscopy Using Convolution Neural Networks

A novel data augmentation algorithm is introduced to improve the model generalization and performance, reduce the cost of data annotation, and achieve robust, high-precision segmentation in the 2D X-ray image of Guidewire tip detection.