Classification and Diagnosis of Lymphoma’s Histopathological Images Using Transfer Learning

Current cancer diagnosis procedure requires expert knowledge and is time-consuming, which raises the need to build an accurate support system for lymphoma identification classification. Many studies have shown promising results using Machine Learning and, recently, Deep detect malignancy in cells. However, diversity complexity of morphological structure make it a challenging classification problem. In literature, many attempts were made classify up four simple types lymphoma. This paper presents approach reliable model capable diagnosing seven different categories rare aggressive These Lymphoma are Classical Hodgkin Lymphoma, Nodular Predominant, Burkitt Follicular Mantle Large B-Cell T-Cell Lymphoma. Our proposed uses Residual Neural Networks, ResNet50, with Transfer lymphoma’s detection The used validated according performance evaluation metrics: Accuracy, precision, recall, F-score, kappa score multi-classes. algorithms tested, on 323 images 224 × pixels resolution. show that our can predict correct subtype accuracy 91.6%.