Detection and Classification of Leukaemia Cells Leukemia is one of the many types of cancers. Leukemia is caused in the white blood cells near the bone marrow region of our body. In this the WBCs which get infected turns blue. Like any other cancer in this also the cell divides itself at the faster pace. Most human cancers are characterized by the aberrant expression of normal and/or mutated genes, and natural selection acts on cancer cells to cause a loss of growth control, angiogenesis, invasion, and metastasis. Even when it is not required they multiply causing a tumor. Detected and treated at an early stage of leukemia saves a lot of lives. The aim of this research is to automate the detection of leukemia cells. In the scientific language, the leukemia cells are known as the blast cells. There are two types of acute leukemia, Acute Lymphoblastic Leukemia (ALL) and Acute Myeloid Leukemia (AML). This thesis focuses on Acute Lymphoblastic Leukemia (ALL). Generally, the process of detection and classification is done manually taking up to five days. The motivation behind this research is to improve the diagnostic process by automating it and reducing its time span to five days to few hours. Nowadays, medical imaging is one of the fastest growing fields in medicine, clinical settings and research and development (RD). Image processing in medical field is becoming a subject of prime focus due to its tremendous potential for the public health sector and the scientific community in general. In particular, imaging applications are emerging as a new opportunity as an innovation at the meeting point between medicine and the computer science. Many software and research groups focus on the development of image processing applications for medical images, for example to improve low resolution photographic images and produce effective high quality images. There is no terrifying disease than cancer nowadays. It is often seen as untreatable, un curable and a very painful disease. Leukemia detection helps in detecting blood cancer using two basic modules of image processing i.e. Image segmentation and feature extraction. After these two modules, we use two techniques of neural network i.e. feed forward network and RBFNN for the detection purposes. We compare the accuracy percentage in both of them. The technique with best accuracy percentage is recorded as the more efficient technique. More than 310,000 Americans are living with leukemia. Every day 143 Americansare detectedwith leukemia and 66 lose the fight.  A brief overview of leukemia and a conceptual analysis of the main methods used for the detection and classification of leukemia cells facilitating Artificial Intelligence, Cellular Automata and Neural Networks are discussed below. Cancer has become a data-intensive range of investigation, with growing amount of changes in data collection technologies and methodologies. In 1895, Wilhelm Roentgen discovered that X-ray tubes, utilized widely for imaging bones and then for giving a variety of circumstances. The technicians who ran the radiograph machines and many exposed patients were found with skin tumors and leukemia. Accurate diagnosis and sorting of blast cells is an tremendously valuable necessity for the detailed diagnosis of leukemia and has a optimistic impact on treatment and prognosis. BLOOD Blood is important part of human life. An average human body is around 70 liters of liquid from which five liters is blood. Biologically, blood is vital for preserving homeostasis that is keeping the bodyâ€™s position stable. This discusses to hydration, temperature regulation and ion concentration. a) Transfer of nutrients from the digestive system to wholly parts of the body. b) Transportation of oxygen from the lungs to all parts of a body. c) Transportation of carbon dioxide from all parts of the body to the lungs. d) Transportation of waste products from cells to the external environment, especially via the kidneys. e) Keeping an ongoing discussion of it is mechanisms with tissue fluids and keeping electrolyte balance. f) Defending the body against attack from foreign viruses through the white blood cells and antibodies. g) Shielding the body against injury or illness using the provocative response. h) Preventing serious hemorrhage by the clotting process. Blood has four main fundamentals to ensure it fulfills its functions, shown in Table 1.1 Table 1.1 (a): Major elements of Blood (Red blood cells, White blood cells, Platelets). Table 1.1(b): Major elements of Blood (Plasma). 1.1.2 WHITE BLOOD CELLS A white blood cell is superior to a red blood cell. White blood cell arrangement and concentration in the blood gives appreciated information and plays a crucial role in the diagnosis of different diseases. White blood cells fall into five categories: Neutrophil, Eosinophil, Basophil, Monocyte and Lymphocyte, shown in the Table 1.2. These cells afford the greatest defense against infections, and their discrete concentrations can help authorities to distinguish between the presences or not of severe pathologies. Types of blood cells are discussed in the following table Table 1.2 : White Blood Cells (Neutrophil, Eosinophil, Basophil, monocyte, lymphocyte) 1.1.3 TYPES OF LEUKEMIA Leukemia is a sickness of unidentified cause where the bone marrow produces huge numbers of irregular cells white blood cells that stop increasing before maturity. There are four main types of leukemia, namely Acute Lymphoblastic Leukemia (ALL), Acute Myeloid Leukemia (AML), which is used as a case study in the thesis, Chronic Lymphocytic Leukemia (CLL) and Chronic Myeloid Leukemia (CML). Most commonly, acute leukemia patients are discussed to specialist units for evaluation. Treatment is based on chemotherapy through the veins, lasting four to six months, which also kills normal body cells. Leukemia can be identified by blood tests while a bone marrow test assists to choose on the best choice of treatment. Table 1.3 includes the types of main leukemia. Table 1.3 : Types of Leukemia Table below shows the UK Leukemia case statistics for males and females in 2007, revealing that the survival rate has increased from 2001 to 2006. The diagnosis and the medical treatment have improved significantly as shown in Figure 2.1. Automated detecting can contribute to the early diagnosis of patients and survival rates are expected to increase in the future. Table 1.4: Leukemia cases in UK for 2007 Figure 1.1: Leukemia 10-year relative survival rates 1.1.4 FLOW CHART OF THE PEOPLE ADMITTED IN THE HOSPITAL Figure 1.2 shows the steps that are essential to be taken by a hematologist in order to identify a patient with acute leukemia. Table 1.5 provides a more thorough explanation of the individual steps in Figure 1.2 NO YES NOYES Table 1.5 : Analytical description of each step in Figure 1.2
Write something about yourself. No need to be fancy, just an overview.