Automated Hematology: Blood Cells Detection and Classification Using Deep Learning Algorithms

A comprehensive assessment of blood cell components serves as a pivotal step in recognizing specific health issues in individuals. The analysis of life-threatening conditions like myeloma and thrombocytopenia underscores the quantification of platelets, erythrocytes (RBCs), and leukocytes (WBCs). Conventional methods, comprising automated analyzers, and manual counting were frequently employed, but they are laborious, lengthy, costly, and require the skills of numerous medical professionals. To mitigate the aforementioned slow and expensive techniques, this study utilizes a deep learning-driven system for object recognition and categorization, employing the You Only Look Once (YOLO) algorithm to count blood cells. The Tiny YOLO model was trained on the BCCD dataset, utilizing a customized architecture designed specifically for detecting the different components of blood cells. Following this, a Convolutional Neural Network (CNN) was utilized to classify the identified WBCs.