Sonam Tyagi

Work place: School of Computing, Graphic Era Hill University, Haldwani, India



Research Interests: Deep Learning, Image Processing, Machine Learning


Sonam Tyagi received her M.Tech in ECE from Kamla Nehru Institute of Technology, Sultanpur, India, in 2016. Currently, she is working as an Assistant Professor in the Department of CSE at Graphic Era Hill University, Haldwani Campus, Uttarakhand, India. Her research interests include Medical Image Processing, Deep Learning, Machine Learning, and Healthcare.

Author Articles
Advances in Medical Imaging: Using Convolutional Neural Networks for White Blood Cell Identification

By Ishwari Singh Rajput Sonam Tyagi Aditya Gupta Vibha Jain

DOI:, Pub. Date: 8 Feb. 2024

White blood cells (WBC) perform a vital function within the immune system by actively protecting the body from a wide range of diseases and foreign substances. Diverse types of WBCs exist, including neutrophils, lymphocytes, eosinophils, and monocytes, each possessing distinct roles within the immune response. Neutrophils are typically the initial immune cells to mobilize in response to infections and inflammation, exhibiting a rapid and robust reaction. Conversely, lymphocytes play a pivotal role in the recognition and targeted elimination of pathogens. Nevertheless, identifying and classifying WBCs poses significant challenges and demands considerable time, even for seasoned medical practitioners. The process of manual classification is frequently characterized by subjectivity and is susceptible to errors, thereby potentially compromising the precision of both diagnosis and treatment. In response to this challenge, scholars have devised deep learning methodologies that can automate the process of WBC classification, thereby enhancing its precision. This study employs a convolutional neural network (CNN) to classify WBCs based on imaging data. The CNN underwent training using a substantial dataset comprising body cell images. This training facilitated the acquisition of discerning characteristics specific to various WBC types, thereby enabling accurate classification. The methodology was evaluated within a simulated environment, yielding encouraging outcomes. The approach that was proposed successfully achieved an average accuracy rate of 98.33% in the classification of WBCs. This outcome serves as evidence of deep learning techniques enhancing the speed and accuracy of WBC classification.

[...] Read more.
Other Articles