Author: Bassi, Pedro R. A. S.; Attux, Romis
Title: A Deep Convolutional Neural Network for COVID-19 Detection Using Chest X-Rays Cord-id: qootuzu7 Document date: 2020_4_30
ID: qootuzu7
Snippet: We present an image classifier based on the CheXNet and a transfer learning stage to classify chest X-Ray images according to three labels: COVID-19, viral pneumonia and normal. CheXNet is a DenseNet121 that has been trained twice, firstly on ImageNet and then, for classification of pneumonia and other 13 chest diseases, over a large chest X-Ray database (ChestX- ray14). The proposed network reached a test accuracy of 97.8% and, for the COVID-19 class, of 98.3%. In order to clarify the modus ope
Document: We present an image classifier based on the CheXNet and a transfer learning stage to classify chest X-Ray images according to three labels: COVID-19, viral pneumonia and normal. CheXNet is a DenseNet121 that has been trained twice, firstly on ImageNet and then, for classification of pneumonia and other 13 chest diseases, over a large chest X-Ray database (ChestX- ray14). The proposed network reached a test accuracy of 97.8% and, for the COVID-19 class, of 98.3%. In order to clarify the modus operandi of the network, we used Layer Wise Relevance Propagation (LRP) to generate heat maps, indicating an analytical path for future research on diagnosis.
Search related documents:
Co phrase search for related documents- acute respiratory and lung healthy: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34
- acute respiratory and lung image: 1, 2, 3, 4, 5, 6, 7
- acute respiratory and lung presence: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51
- acute respiratory and machine expert: 1, 2
Co phrase search for related documents, hyperlinks ordered by date