A Survey on Blood Image Diseases Detection Using Deep Learning

Abstract

Blood disease detection and diagnosis using blood cells images is an interesting and active research area in both the computer and medical fields. There are many techniques developed to examine blood samples to detect leukemia disease, these techniques are the traditional techniques and the deep learning (DL) technique. This article presents a survey on the different traditional techniques and DL approaches that have been employed in blood disease diagnosis based on blood cells images and to compare between the two approaches in quality of assessment, accuracy, cost and speed. This article covers 19 studies, 11 of these studies were in traditional techniques which used image processing and machine learning (ML) algorithms such as K-means, K-nearest neighbor (KNN), Naïve Bayes, Support Vector Machine (SVM), and 8 studies in advanced techniques which used DL, particularly Convolutional Neural Networks (CNNs) which is the most widely used in the field of blood image diseases detection since it is highly accurate, fast, and has the least cost. In addition, it analyzes a number of recent works that have been introduced in the field including the size of the dataset, the used methodologies, the obtained results, etc. Finally, based on the conducted study, it can be concluded that the proposed system CNN was achieving huge successes in the field whether regarding features extraction or classification task, time, accuracy, and had a lower cost in the detection of leukemia diseases.

COVID-19 cough sound symptoms classification from scalogram image representation using deep learning models

“…The introduced model includes two main components: The first component is the feature extraction, and it transforms sound to image based on scalogram, while the second component is the feature extraction and classification model based on the DL models (GoogleNet, ResNet18, ResNet50, ResNet101, MobileNetv2, and NasNetmobile). GoogleNet, ResNet, MobileNet, and NasNet are amongst the most widely used DL transfer learning models [ [34] , [35] , [36] ]. The proposed model used DL models for feature extraction and classification in the training, validation, and testing stages.…”

“…5 . It contains 22 convolutional layers [ 34 , 35 ]. GoogLeNet has inception layers, each of which conducts a particular method of convolution and then concatenates the filters together for the next layer [ 40 ].…”

A Novel Hybrid Classifier Model for Leukemia Prediction Using Machine Learning Techniques

“…The classification of leukaemia is done to summarize into four general kinds called Chronic Lymphoid or Lymphoblastic Leukaemia (CLL), Acute Lymphoid or Lymphoblastic Leukaemia (ALL), Chronic Myeloid Leukaemia (CML), and Acute Myeloid Leukaemia (AML) [8]. The most general leukaemia is ALL, which contributes 12% of complete cases and the prevalence of ALL between the age group 1 to 12 years is 80% [9]. The immature lymphocyte cells in the smear blood image are characterized by the ALL progressing quickly.…”

“…On the other hand, during childhood and adolescence, it has an incidence of around 1 in 1, 50,000. It is prevalent mostly among the patients of 40 years age group [9]. About 25% of patients' reports are founded in the plain bone and joint at the AML inception.…”

Categorizing white blood cells by utilizing deep features of proposed 4B-AdditionNet-based CNN network with ant colony optimization

“…It makes up about 7% of an adult's body weight [1]. It is composed of 55% plasma which allows it to flow freely throughout the body using blood vessels [2]. Centered on the color, size, texture, shape, and composition, the cellular components of blood are separated into three cell types i.e., erythrocytes (red blood cells or RBCs) [3], leukocytes (WBCs) [4,5], and thrombocytes (platelets) [6].…”