Radiologist-Level COVID-19 Detection Using CT Scans with Detail-Oriented Capsule Networks
Authors:
Aryan Mobiny,
Pietro Antonio Cicalese,
Samira Zare,
Pengyu Yuan,
Mohammadsajad Abavisani,
Carol C. Wu,
Jitesh Ahuja,
Patricia M. de Groot,
Hien Van Nguyen
Abstract:
Radiographic images offer an alternative method for the rapid screening and monitoring of Coronavirus Disease 2019 (COVID-19) patients. This approach is limited by the shortage of radiology experts who can provide a timely interpretation of these images. Motivated by this challenge, our paper proposes a novel learning architecture, called Detail-Oriented Capsule Networks (DECAPS), for the automati…
▽ More
Radiographic images offer an alternative method for the rapid screening and monitoring of Coronavirus Disease 2019 (COVID-19) patients. This approach is limited by the shortage of radiology experts who can provide a timely interpretation of these images. Motivated by this challenge, our paper proposes a novel learning architecture, called Detail-Oriented Capsule Networks (DECAPS), for the automatic diagnosis of COVID-19 from Computed Tomography (CT) scans. Our network combines the strength of Capsule Networks with several architecture improvements meant to boost classification accuracies. First, DECAPS uses an Inverted Dynamic Routing mechanism which increases model stability by preventing the passage of information from non-descriptive regions. Second, DECAPS employs a Peekaboo training procedure which uses a two-stage patch crop and drop strategy to encourage the network to generate activation maps for every target concept. The network then uses the activation maps to focus on regions of interest and combines both coarse and fine-grained representations of the data. Finally, we use a data augmentation method based on conditional generative adversarial networks to deal with the issue of data scarcity. Our model achieves 84.3% precision, 91.5% recall, and 96.1% area under the ROC curve, significantly outperforming state-of-the-art methods. We compare the performance of the DECAPS model with three experienced, well-trained thoracic radiologists and show that the architecture significantly outperforms them. While further studies on larger datasets are required to confirm this finding, our results imply that architectures like DECAPS can be used to assist radiologists in the CT scan mediated diagnosis of COVID-19.
△ Less
Submitted 15 April, 2020;
originally announced April 2020.
