Extracting Possibly Representative COVID-19 Biomarkers from X-ray Images with Deep Learning Approach and Image Data Related to Pulmonary Diseases.

Ioannis D ApostolopoulosSokratis I AznaouridisMpesiana A Tzani

Published in: Journal of medical and biological engineering (2020)

The results suggest that training CNNs from scratch may reveal vital biomarkers related but not limited to the COVID-19 disease, while the top classification accuracy suggests further examination of the X-ray imaging potential.

Keyphrases

deep learning
high resolution
coronavirus disease
sars cov
convolutional neural network
artificial intelligence
machine learning
dual energy
big data
magnetic resonance imaging
computed tomography
electronic health record
cross sectional
single cell
climate change
human health
optical coherence tomography