Leukemia Blood Cell Image Classification Using Convolutional Neural Network

Acute myeloid leukemia is a type of malignant blood cell cancer that can affect both children and adults.There are 60,140 people were expected to be diagnosed with Leukemia in 2016, according to the Leukemia and Lymphoma Society.In order to get the most effective treatment, the patient needs early diagnosis.Therefore we need to have a support system of early diagnosis to guide treatment for patients with acute leukemia as soon as possible.In this paper, the authors propose a Convolutional Neural Network (CNN) based method to distinguish normal and abnormal blood cell images.The proposed method achieves an accuracy up to 96.6% with the dataset including 1188 blood cell images.