Detection of WBC Cancer using Image Processing

Abstract

White Blood Cells (WBC) Cancer detection is a tedious task because cancer diseases worsen rapidly in a short period of time. The Manual diagnosing systems used today to detect WBC cancers are Lumbar Puncture, Bone Marrow Biopsy and Lymph Node Biopsy. These systems are not only time consuming and expensive, but might be inaccurate sometimes which leads to a misdiagnosis in most cases posing a life-threatening situation for patients. To avoid these situations, this survey paper proposes a method by which an automated system is developed and designed to ease the detection of cancer disease in a short period of time and also which is cost efficient. The aim of the system is to produce a highly accurate and promising results in diagnosing the cancer using digital image processing method of Machine Learning. And an automatic model is developed to work without any requirement of the lab technicians to analyze the result. A robust segmentation, clustering and deep learning techniques like Multi-Layer Perceptron will be used and for classification Support Vector Machine (SVM) will be used to achieve accurate results on the bone-marrow images by training the system. The proposed system will detect the WBC cancer cells with the less amount of dataset by undergoing several processes.