COVID-19 PREDICTION THROUGH CHEST X-RAY IMAGE DATASETS USING DEEP LEARNING
12th International Conference on Computing Communication and Networking Technologies, ICCCNT 2021
; 2021.
Article
in English
| Scopus | ID: covidwho-1752380
ABSTRACT
The work aims at the prediction and analysis of COVID-19 from Chest X-Ray scan images using Pre-trained Deep Convolutional Neural Network models. Analysis is carried out using two open-source datasets, to identify and differentiate between the Chest X-Ray scans of non COVID person and COVID-19 affected person. A baseline model using LeNet-5 is implemented using the initial dataset collected, which gave 98.57 % accuracy. Further, pre-trained models such as AlexNet, ResNet 50, Inception V3, VGG16, VGG19 and Xception are used for COVID prediction and carryout comparative analysis. Using the performance measures viz. Accuracy, Confusion Matrix and ROC Curves, the result of study shows that for the first dataset used for analysis, Xception and for the second dataset Inception architectures respectively are most suitable for the prediction of COVID-19. © 2021 IEEE.
Full text:
Available
Collection:
Databases of international organizations
Database:
Scopus
Type of study:
Prognostic study
Language:
English
Journal:
12th International Conference on Computing Communication and Networking Technologies, ICCCNT 2021
Year:
2021
Document Type:
Article
Similar
MEDLINE
...
LILACS
LIS