Assessment of Critical Feeding Tube Malpositions on Radiographs Using Deep Learning
| Autor: | Paras Lakhani, Varun Danda, Adam E. Flanders, Richard Gorniak, Varun Singh |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2019 |
| Předmět: |
Radiography
Abdominal medicine.medical_specialty Artificial ntelligence Radiography Chest radiography Convolutional neural network Article 030218 nuclear medicine & medical imaging 03 medical and health sciences 0302 clinical medicine Enteral Nutrition Machine learning medicine Image Processing Computer-Assisted Humans Radiology Nuclear Medicine and imaging Feeding tube Radiological and Ultrasound Technology Receiver operating characteristic Medical Errors business.industry Deep learning Area under the curve Computer Science Applications Rapid identification Tube placement Radiography Thoracic Radiology Artificial intelligence Neural Networks Computer business 030217 neurology & neurosurgery |
| Zdroj: | Journal of Digital Imaging |
| ISSN: | 1618-727X 0897-1889 |
| Popis: | Assess the efficacy of deep convolutional neural networks (DCNNs) in detection of critical enteric feeding tube malpositions on radiographs. 5475 de-identified HIPAA compliant frontal view chest and abdominal radiographs were obtained, consisting of 174 x-rays of bronchial insertions and 5301 non-critical radiographs, including normal course, normal chest, and normal abdominal x-rays. The ground-truth classification for enteric feeding tube placement was performed by two board-certified radiologists. Untrained and pretrained deep convolutional neural network models for Inception V3, ResNet50, and DenseNet 121 were each employed. The radiographs were fed into each deep convolutional neural network, which included untrained and pretrained models. The Tensorflow framework was used for Inception V3, ResNet50, and DenseNet. Images were split into training (4745), validation (630), and test (100). Both real-time and preprocessing image augmentation strategies were performed. Receiver operating characteristic (ROC) and area under the curve (AUC) on the test data were used to assess the models. Statistical differences among the AUCs were obtained. p |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink