Résumé:
Determining cancer and its type is a very difficult task that requires high medical expertise and skills. With the development of image classification techniques, deep learning strategies have occupied the first positions in many medical image classification systems as part of computer aide decision (CAD).
The aim of this study is to accurately classify lymphoma subtypes using deep learning. A deep learning framework has been proposed to classify three types of lymphomas as follicular lymphoma (FL), chronic lymphocytic lymphoma (CLL) and Mantle Cell Lymphoma (MCL) by following pretrained CNN models (Transfer learning) such as Resnet and VGG and based on the available dataset from the National Institute on Aging (NIA). The data Patching was implemented for the first step of data processing, where the achieved results show that the proposed models were able to achieve better results compared to CNN built from scratch
