Presenter:

Prof. My Abdelouahed SABRI

• Professor
• Computer Science Department
• Faculty of sciences, Dhar El Mahraz
• University Sidi Mohamed Ben Abdellah
• Morocco
• WebSite: https://sites.google.com/a/usmba.ac.ma/sabri/
• Google Scholar https://scholar.google.com/citations?user=bPJDulYAAAAJ

Abstract:

Computer-Aided Diagnosis systems have been widely applied in medical image classification. They use the same process as the Content-Based Information Retrieval (CBIR); the offline phase which consists of creating a classification model and the inline phase to predict or classify unknown images using the elaborated classification model. The classification model is elaborated based on extracted features from the training dataset and evaluated on the validation dataset. Features used are generally based on shape, color, and/or texture information. Features extraction step can be avoided using deep learning methods but remains less efficient when dealing with small datasets or high-resolution images. To better understand the image classification process, it would be good to start from the beginning. Going through the CBIR, towards the development of powerful models for the classification of medical images using Machine Learning algorithms. Our examples of applications will be based on the classification of skin cancer images from the well-known dataset used in the ISIC2017 challenge.

Workshop Program:

To be developed

References