Depot Institutionnel de l'UMBB >
Mémoires de Master 2 >
Institut de Génie Electrique et d'Electronique >
Computer >
Veuillez utiliser cette adresse pour citer ce document :
http://dlibrary.univ-boumerdes.dz:8080/handle/123456789/11887
|
Titre: | GPU-based COVID-19 and other lung infections detector from chest X-rays using deep convolutional neural networks. |
Auteur(s): | Rezig, Warda Bouazza, Manel Namane, Rachid (Supervisor) |
Mots-clés: | COVID-19 detector Convolutional Neural Networks |
Date de publication: | 2021 |
Résumé: | The world has been tormented last year by a deadly virus outbreak that had ravaged, and still is, many humans’ lives across the globe. Fighting against this coronavirus disease requires effective and fast screening methods. This study aims to leverage deep learning techniques to build a Deep Convolutional Neural Network to detect COVID-19 among Normal and other lung infections namely; Pneumonia, and Lung opacity using chest X-Ray images.
Publicly available X-ray images (3388 Healthy, 1345 Pneumonia, 3388 Lung Opacity and 3388 confirmed COVID-19) are used in a four-class classifier that distinguishes COVID-19 among the other classes aiming at assisting the healthcare community allowing faster screening and hence higher rates of contagion control. To ensure high accuracy levels, a CNN-based model is proposed through an incremental approach as well as the use of pre-trained models. An NVIDIA GeForce GTX 1060 Graphics Processing Unit (GPU) is used to accelerate the detection for an optimal training time. The obtained overall accuracies for these models ranged from 75% to more than 93% with up to 97.76% COVID-19 detection indicating the applicability of deep learning methods in the clinical diagnosis of the virus. |
Description: | 53 p. |
URI/URL: | http://dlibrary.univ-boumerdes.dz:8080/handle/123456789/11887 |
Collection(s) : | Computer
|
Fichier(s) constituant ce document :
|
Tous les documents dans DSpace sont protégés par copyright, avec tous droits réservés.
|