MOHAMED ABDERAOUF Moustari

User profile picture

MOHAMED ABDERAOUF Moustari

محمد عبد الرؤوف مستاري

mohamedabderaouf.moustari@univ-msila.dz

0558221125

  • Departement of ELECTRONICS
  • Faculty of Technology
  • Grade PHd
Send Mail

About Me

Domain

Science et Technologies

Filiere

Electronique

Location

Biskra, Biskra

Biskra, ALGERIA

Code RFIDE

  • 1998-04-19 00:00:00

  • MOHAMED ABDERAOUF Moustari birthday

  • 2024-06-27
  • 2024-06-27

    Two-stage deep learning classification for diabetic retinopathy using gradient weighted class activation mapping

    The fundus images of patients with Diabetic Retinopathy (DR) often display numerous lesions scattered across the retina. Current methods typically utilize the entire image for network learning, which has limitations since DR abnormalities are usually localized. Training Convolutional Neural Networks (CNNs) on global images can be challenging due to excessive noise. Therefore, it's crucial to enhance the visibility of important regions and focus the recognition system on them to improve accuracy. This study investigates the task of classifying the severity of diabetic retinopathy in eye fundus images by employing appropriate preprocessing techniques to enhance image quality. We propose a novel two-branch attention-guided convolutional neural network (AG-CNN) with initial image preprocessing to address these issues. The AG-CNN initially establishes overall attention to the entire image with the global branch and then incorporates a local branch to compensate for any lost discriminative cues. We conduct extensive experiments using the APTOS 2019 DR dataset. Our baseline model, DenseNet-121, achieves average accuracy/AUC values of 0.9746/0.995, respectively. Upon integrating the local branch, the AG-CNN improves the average accuracy/AUC to 0.9848/0.998, representing a significant advancement in state-of-the-art performance within the field.

    Citation

    Mohamed Abderaouf Moustari , , (2024-06-27), Two-stage deep learning classification for diabetic retinopathy using gradient weighted class activation mapping, Automatika: Journal for Control, Measurement, Electronics, Computing and Communications, Vol:65, Issue:3, pages:1284-1299, Taylor and Francis Ltd.

    See publication
  • 2022-11-26
  • 2022-11-26

    Enhancement of diabetic retinopathy classification using attention guided convolution neural network

    Damage to the retina from diabetes can lead to permanent vision loss due to a condition known as diabetic retinopathy. In order to avoid this, it is essential to diagnose this disease early. To address these problems, this paper proposes a two-branch Grad-CAM attention-guided convolution neural network (AG-CNN) with initial CLAHE image preprocessing. The AG-CNN first builds a general attention to the entire image with the global branch, in order to further concentrate the system's attention on the localized areas of the problems, the system isolate the important regions (ROIs) of the global image and then feeds them to a local branch. This extensive experiment is based on the APTOS 2019 DR dataset. In order to start, we offer a solid global baseline that, using DenseNet-121 as a starting point, produced average accuracy/AUC values of 0.9746/0.995, respectively. The average accuracy and AUC of the AG-CNN are increased to 0.9848 and 0.998, respectively, after creating the local branch. which represents a new state-of-the-art in the field.

    Citation

    Mohamed Abderaouf Moustari , BRIK Youcef , BILAL Attallah , ,(2022-11-26), Enhancement of diabetic retinopathy classification using attention guided convolution neural network,ICATEEE2022,M'sila, Algeria

    See publication
  • 2021
  • 2021

    Diabetic Retinopathy Classification from Retinal Images using an “Attention Guided” Deep Learning Approach

    This workshop presentation looks at the problem of classifying the severity of diabetic retinopathy based on pictures of the inside of the eye. Only the global image is typically used as the network learning input in existing approaches. This method isn't perfect because abnormalities in the eye fundus tend to be in small areas, and training CNNs with a global image may be affected by (too many) irrelevant noisy areas. In this paper, we propose a two-branch attention-guided convolution neural network to overcome the aforementioned issues (AG-CNN). In order to further focus the system's attention on the localized areas of the problems present in the global image, the AG-CNN first builds a general attention to the entire image with the global branch. It then constructs the local branch and incorporates a global branch to make up for the lost discriminative cues by the local branch. Specifically, utilizing global pictures, we first train a global CNN branch. Following that, we infer a mask to crop a discriminative region from the global image using the attention heat map produced by the global branch as our guidance. The local area is utilized to train a local CNN branch that produces the original global image's severity level. The APTOS 2019 DR dataset serves as the basis for this experiment. After constructing the global/local branch, AG-CNN gave the average accuracy/AUC of 0.9629/0.991, respectively.

    Citation

    Mohamed Abderaouf Moustari , ,(2021), Diabetic Retinopathy Classification from Retinal Images using an “Attention Guided” Deep Learning Approach,Journée Doctoral en Electronique (JDE'2022),Universtiy de M'sila

    See publication