DC Field | Value | Language |
dc.contributor.author | BELLAOUEDJ, ISmail | - |
dc.date.accessioned | 2022-11-14T07:29:25Z | - |
dc.date.available | 2022-11-14T07:29:25Z | - |
dc.date.issued | 2022 | - |
dc.identifier.uri | https://repository.esi-sba.dz/jspui/handle/123456789/380 | - |
dc.description | Encadreur : Mr HADJILA Fethallah Co-Encadreur : Mr AMAR BENSABER Djamel | en_US |
dc.description.abstract | Currently, several studies are discussing the detection of lung diseases by analyzing medical
images using deep learning. Deep learning is also a valuable aid to experts in the
interpretation of medical images. Heuristics such as transfer learning (a deep learning
method where a model developed for a given task is reused as a starting point for another
model that targets a second task) is emerging. These approaches (based on pre-trained
models) are used as the starting point for computer vision tasks and can offer considerable
gains to many problems.
In this project, we propose Convolutional Neural Networks (CNN) models to detect the
presence of COVID-19 in X-ray images. As an additional contribution, we show the layers
of the network, i.e. the areas of the chest X-ray that the model considered for generating
the prediction; this can ultimately be presented as a suggestion for the radiologist in the
suspect area.
***
Actuellement, plusieurs ´etudes discutent la d´etection des maladies pulmonaires en
analysant images m´edicales en exploitant l’apprentissage profond (deep learning).Ce
dernier constitue aussi une aide pr´ecieuse aux experts pour l’interpr´etation des images
m´edicales. Des heuristiques telles que l’apprentissage par transfert(une m´ethode
d’apprentissage profond o`u un mod`ele d´evelopp´e pour une tˆache est r´eutilis´e comme point
de d´epart d’un mod`ele sur une seconde tˆache) est en train d’´emerger. Ces approches `a base
de mod`eles pr´e-entraˆın´es sont utilis´es comme le point de d´epart des tˆaches de vision par
ordinateur et peuvent leur assurer d’ ´enormes gains. Nous proposons dans ce projet des
mod`eles de type R´eseau neuronal `a convolution pour d´etecter la pr´esence de COVID-19
dans les images rayons X. Comme contribution suppl´ementaire, nous montrons les couches
d’activation du r´eseau, c’est-`a-dire les zones de la radiographie pulmonaire que le mod-
`ele a consid´er´e pour g´en´erer la pr´ediction, Cela peut repr´esenter une suggestion pour le
radiologue de la zone suspecte. | en_US |
dc.language.iso | en | en_US |
dc.subject | Deep Learning | en_US |
dc.subject | Convolutional Neural Networks | en_US |
dc.subject | Covid-19 Detection | en_US |
dc.subject | Transfer Learning Transfer | en_US |
dc.title | D´etection du COVID-19 par r´eseaux convolutionnels et `a l’aide des images radiologiques (X-rays) | en_US |
dc.type | Thesis | en_US |
Appears in Collections: | Ingénieur
|