Doaa Al-Desouky
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Doaa Al-Desouky has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Ehab Essa
Department of Computer Science, Faculty of Computers and Information, Mansoura University, Mansoura, Egypt
1 shared publication - M. Z. Rashad
Faculty of Computers and Information, Computer Science Dept. Mansoura University, Egypt
1 shared publication - Sherif E. Hussein
Tobacco Control Research Group, Department for Health, University of Bath, Bath, UK
1 shared publication
Latest Publications
- Open Access
- 1 December 2019
- Published by Egyptian Knowledge Bank in Mansoura Journal for Computer and Information Sciences
- Vol. 15 (2) , 1-9
Automatic Multiple Sclerosis lesion segmentation using Patch-wise R-CNN
Multiple sclerosis (MS) could be considered one of the most severe neurological diseases, which can cause damage to the central nervous system. Because of the regular change in size, location and anatomical variation of MS lesions, it is achallenge to accurately identify, characterize and quantify MS lesions on magnetic resonance imaging (MRI). Therefore, MS lesion segmentation and detection become an active point of research. Recently, deep neural networks (DNN) have seen a rapid advance in various medical image analysis fields, i.e., image registration, image segmentation, lesion detection, and shape modeling. Furthermore, convolution neural networks (CNN) have gained popularity in medical imaging, especially in brain imaging.In this study, an automated technique is proposed to segment MS lesions in MRI. This technique depends on a 3D patchwise region-based convolution neural network (R-CNN) for MS lesion segmentation in T2-w and FLAIR.The proposed method is evaluated using the public MICCAI2008 MS lesion segmentation data set, which iscompared to other MS lesion segmentation tools.
- 1 December 2019
- Published by Egyptian Knowledge Bank in Mansoura Journal for Computer and Information Sciences
- Vol. 15 (2) , 1-9
Open AccessAutomatic Multiple Sclerosis lesion segmentation using Patch-wise R-CNN
Multiple sclerosis (MS) could be considered one of the most severe neurological diseases, which can cause damage to the central nervous system. Because of the regular change in size, location and anatomical variation of MS lesions, it is achallenge to accurately identify, characterize and quantify MS lesions on magnetic resonance imaging (MRI). Therefore, MS lesion segmentation and detection become an active point of research. Recently, deep neural networks (DNN) have seen a rapid advance in various medical image analysis fields, i.e., image registration, image segmentation, lesion detection, and shape modeling. Furthermore, convolution neural networks (CNN) have gained popularity in medical imaging, especially in brain imaging.In this study, an automated technique is proposed to segment MS lesions in MRI. This technique depends on a 3D patchwise region-based convolution neural network (R-CNN) for MS lesion segmentation in T2-w and FLAIR.The proposed method is evaluated using the public MICCAI2008 MS lesion segmentation data set, which iscompared to other MS lesion segmentation tools.