COVIDNet: An Automatic Architecture for COVID-19 Detection With Deep Learning From Chest X-Ray Images
Autor: | Xiuying Shi, Prayag Tiwari, Lang He, Rui Su, Neeraj Kumar, Pekka Marttinen |
---|---|
Přispěvatelé: | Xi'an University of Posts and Telecommunications, Department of Computer Science, Northwest University, Yan’an University, Asia University Taiwan, Aalto-yliopisto, Aalto University |
Rok vydání: | 2022 |
Předmět: |
Coronavirus disease 2019 (COVID-19)
Computer Networks and Communications Computer science business.industry Deep learning Internet of Things COVID-19 Computer Science Applications Databases Hardware and Architecture Solid modeling Signal Processing X ray image Feature extraction Training Computer vision Artificial intelligence Architecture business Computed tomography Information Systems |
Zdroj: | IEEE Internet of Things Journal. 9:11376-11384 |
ISSN: | 2372-2541 |
Popis: | openaire: EC/H2020/101016775/EU//INTERVENE Up to now, the COVID-19 has been sweeping across all over the world, which has affected individual’s lives in an overwhelming way. To fight efficiently against the COVID-19, radiography and radiology images are used by clinicians in hospitals. This paper presents an integrated framework, named COVIDNet, for classifying COVID-19 patients and healthy controls. Specifically, ResNet (i.e., ResNet-18 and ResNet-50) is adopted as a backbone network to extract the discriminative features first. Second, the spatial pyramid pooling (SPP) layer is adopted to capture the middle-level features from the features of ResNet. To learn the high-level features, the NetVLAD layer is used to aggregate the features representation from middle-level features. Context gating (CG) mechanism is adopted to further learn the high-level features for predicting the COVID-19 patients or not. Finally, extensive experiments are conducted on the collected database, showing the excellent performance of the proposed integrated architecture, with the sensitivity up to 97%, and specificity of 99.5% of the ResNet-18, and with the sensitivity up to 99%, and specificity of 99.4% of the ResNet-50. |
Databáze: | OpenAIRE |
Externí odkaz: |