What Are The Different Types Of Leukemia – HD-OCT testing may be useful in detecting early stage markers of diabetic retinopathy in patients with type 1 diabetes.
Diagnostic Performance of a Support Vector Device for Dermatofluoroscopic Melanoma Detection: Results of a Retrospective Clinical Study of 214 Pigmented Skin Lesions
What Are The Different Types Of Leukemia
Open Access Policy Access Program Guidelines Screening for Special Topics Research Editorial Policy and Publishing Ethics Fees for Processing Articles Fees Ideas
Acute Lymphoblastic Leukemia Vs. Chronic Lymphocytic Leukemia
All articles published by us are immediately available worldwide under an open access license. No special permission is required to reuse all or part of the material published by , including figures and tables. For articles published under the Creative Commons CC BY open access license, any part of the article may be reused without permission, as long as the original article is clearly cited. For more information, please see https:///openaccess.
These papers represent the most advanced research with significant potential for significant impact in the field. The Feature Paper should be a significant original article that involves multiple methods or approaches, provides a perspective for future research directions, and describes possible research applications.
Feature papers are submitted by individual invitation or recommendation from scientific editors and must receive positive feedback from reviewers.
Editor’s Choice articles are based on recommendations from scientific editors of journals from around the world. The editors select a small number of recently published articles in the journal that they believe will be of particular interest to readers or important in the respective research area. The aim is to provide a snapshot of some of the most interesting works published in various research areas.
Do You Know About The Different Types Of Leukemia? Expert Explains About Their Treatment Options
By Nizar Ahmed Nizar Ahmed Scilit Preprints.org Google Scholar, Altug YigitAltug Yigit SciProfiles Scilit Preprints.org Google Scholar, Zerrin IsikZerrin Isik SciProfiles Scilit Preprints.org Google Scholar and Adil AlpkocakAdil Alpkocak ScilitProfiles.
Submission received: June 2, 2019 / Revised: August 22, 2019 / Accepted: August 23, 2019 / Published: August 25, 2019
Leukemia is a deadly cancer and has two main types: Acute and Chronic. Each type has two additional components: lymphoid and myeloid. So, there are a total of four types of leukemia. This study proposes a novel method for diagnosing all subtypes of leukemia from microscopic blood cell images using convolutional neural networks (CNN), which requires a large training data base. Therefore, we also investigate the effects of data enrichment for an increasing number of synthetically trained samples. We used two publicly available leukemia data sources: ALL-IDB and ASH Image Bank. Next, we used seven different image transformation techniques as data enhancement. We designed a CNN architecture capable of recognizing all subtypes of leukemia. In addition, we also explore other well-known machine learning algorithms, such as naive Bayes, support vector machine, k-nearest neighbor, and decision tree. To evaluate our method, we set up a set of experiments and used cross-validation 5. The results obtained from the experiments show that the performance of the CNN model is 88.25% and 81.74% accurate in leukemia versus healthy and multi-classification of all tax types, respectively. Finally, we also show that the CNN model has a better performance than other well-known machine learning algorithms.
Leukemia diagnosis; identification of leukemia subtypes; multi-categorization; microscopic blood cell images; data augmentation; deep learning; convolutional neural network
Five Things To Know About Blood Cancers
Leukemia is an aggressive disease related to white blood cells (WBC) and affects the bone marrow and blood in the human body. This disease can lead to the destruction of a person’s immune system. There are two main types of leukemia, acute and chronic leukemia, which depend on how quickly it progresses. In acute leukemia, infected WBCs do not behave like normal WBCs; while it may function as a normal WBC in chronic leukemia. Therefore, chronic leukemia may be important because it is indistinguishable from normal WBC. In addition, there are two types of leukemia each based on the size and shape of the WBC: Lymphoid and myeloid. In general, there are four types of leukemia as shown in Figure 1, acute lymphocytic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL), and chronic myeloid leukemia (CML) [ 1, 2]. Analyzing the existence of leukemia and their specific types is important for scientists to avoid medical risks and determine the right therapy. Therefore, the use of cognitive methods of diagnosis will facilitate and speed up the detection of leukemia subtypes using blood cell images (ie blood smears).
Microscopic blood tests are considered the most important procedure for leukemia diagnosis [2]. Analysis of blood smears is the most common way to detect leukemia, but it is not the only procedure. Interventional radiology is the technique of choice for diagnosing leukemia. However, radiological techniques, such as percutaneous aspiration, biopsy and catheter drainage, suffer from inherent limitations of imaging sensitivity and resolution of radiographic images [3]. Furthermore, other techniques, such as Molecular Cytogenetics, Long Distance Inverse Polymerase Chain Reaction (LDI-PCR) and Array-based Comparative Genomic Hybridization (aCGH), require a lot of work and time to identify types of leukemia [4]. Because of the time and cost requirements of these procedures, microscopic blood and bone marrow tests are the most common methods for identifying leukemia subtypes.
A machine learning (ML) algorithm will help identify blood cells with leukemia from HEALTHY cells when there is a large training set. The ALL-IDB leukemia image database [5] is one of the databases used by a number of medical researchers [1, 6, 7] as a benchmark. Another leukemia database is from the American Society of Hematology (ASH) and is available online at the website [8]. Thanh et al. [7] used the ASH database to identify AML leukemia in their study. Google is another source of leukemic images without comments, where images are collected randomly from websites. Karthikeyan et al. [9] used microscopic images obtained from Google in their study to identify leukemia, where the authors themselves annotated the images. The successful implementation of machine-based leukemia diagnosis can be built on the use of descriptive imaging data.
Identification of leukemia subtypes from HEALTH samples is a very challenging problem. In the literature, many researchers investigate the binary distinction only within a species rather than HEALTH samples [1, 7, 9, 10, 11, 12]. They achieve very high efficiency, even more than 96% accuracy. Also, Shafique et al. [6] further classified samples with ALL components according to the size of the cell and the nature of its nucleus. However, identifying all subtypes of leukemia is a more challenging task than a simple binary classification [13]. To our knowledge, there is no automatic identification method that captures all leukemia subtypes.
Hairy Cell Leukemia: What Is It, How Common Is It, And More
Several ML algorithms help to differentiate and identify leukemia from microscopic images. For example, Paswan et al. [10] who used support vector machine (SVM) and k-nearest neighbor (k-NN) to classify AML leukemia subtype, achieved 83% accuracy. Patel et al. [1] applied SVM to ALL leukemia subtypes and achieved 93% accuracy. Karthikeyan et al. [9] also used SVM and c-means clustering technique to separate WBC from background and they reached 90% accuracy. Although the use of deep learning (DL) method seems to be more effective, its performance is highly dependent on the quantity and quality of the data set used [6]. Convolutional Neural Network (CNN) is one of the neural network architectures generally used to deal with image classification and registration problems. Shafique et al. [6] used a convolutional neural network (CNN) to identify ALL leukemia subtypes. Their results recorded 99% for binary classification, between ALL according to HEALTH samples, and 96% for further classification of subtypes of ALL alone. Thanh et al. [7] also built a CNN model with five convolutional layers to perform binary classification of ALL leukemia subtypes and achieved 96.6% accuracy. Unfortunately, functional classification in such a neural network requires large training data to learn to identify important objects from every image. However, developing a large training database is a time-consuming and very labor-intensive task. To avoid this problem, we suggest to expand the limited number of samples by image compression. Using an insufficient number of image samples in the training document may lead to overfitting problem [14]. Therefore, most of the researchers in the literature rely on the use of some image transformation techniques to increase the number of training set samples to avoid the overload problem. Patel et al. [1] used median and Wiener filters to remove noise and blur. In the literature, many image transformation techniques have been used, such as image rotation and distortion, histogram equalization, image translation, gray transformation, image shaking, and image transformation [6, 9, 10]. The use of image enhancement allows the DL method to be used, which requires a large number of training datasets.
In this study, we propose a novel method for leukemia diagnosis from microscopic blood images that identifies four types of leukemia (i.e., ALL, AML, CLL, and CML) using CNN architecture theory. To our knowledge, this is the first study to address all four leukemias