Please use this identifier to cite or link to this item: http://archives.univ-biskra.dz/handle/123456789/13255
Title: Magnetic resonance imaging (MRI)/Segmentation methods of MRI images/Convolutional Neural Network for Stroke Lesion Segmentation and Implementation 3
Other Titles: informatique
Authors: kebache, rofaida
Issue Date: 20-Jun-2019
Abstract: Stroke is one of the leading causes of death and disability. Therefore the manual segmentation of it's lesions is time-consuming, automatic segmentation methods of the stroke has recently extracted attentions. One of the successful automatic methods that achieve state of the art is the deep learning (convolutional neural network), this method is used in multiple domains such that image recognition, Medical context, natural language processing etc ... The need for a short prediction time and an accurate segmentation is a challenge to work on that's why all the present and the future works are focusing on the variety architectures witch get the score and to achieve the state of the art. We intent in this work to build a Convoultional neural network for the task of the segmentation,our contribution is that we based our model on the Auto-encoder network by inspiring from their main idea and by using a CNN layers with multispectral MRI images to improve the robustness of our model.
URI: http://archives.univ-biskra.dz/handle/123456789/13255
Appears in Collections:Faculté des Sciences Exactes et des Science de la Nature et de la vie (FSESNV)

Files in This Item:
File Description SizeFormat 
kebache_roufaida.pdf1,7 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.