Computer-aided Detection of White Blood Cells Using Geometric Features and Color

White blood cells make up around 1% of our blood, playing a major role in our immune system, fighting foreign organisms and protecting our internal systems. Five different types of leukocytes exist: monocytes, neutrophils, lymphocytes, eosinophils and basophils. In this work, we present a computer-based technique that relies on feature segmentation and extraction in order to efficiently classify white blood cells. Eight features related to the geometry and color of these cells were extracted from 253 images and fed into the random forest classifier. An accuracy of 86% and a precision of 88% were obtained on the testing set. The results indicate that this technique may be used to classify various types of white blood cells.