Article Preview
Top1. Introduction
In recent years, feature selection in big data becomes tremendously important in several life sectors such as medicine, engineering and science (Aziz et al., 2013a,b, 2012;; Hassanien et al., 2019a,b, 2017, 2014a,b, 2015; Ahmed et al., 2020; Mallek et al., 2020; Asad et al., 2012, 2014a,b,c,d; Jothi et al. 2013; Sayed et al., 2019, 2020). One of the critical issues in analyzing biomedical data is the curse of dimensionality especially the data with few samples and presented in high dimensional feature space (Inbarani et al., 2020, 2018, 2014a,b,c,d, 2016, 2015a,b, 2012, 2013; Inbarani and Banu, 2012; Muliawaty et al., 2019). Most of the medical tissues are obviously 3D images, so that irrelevant features not only lead to insufficient classification accuracy, but also reduces the processing speed or computational time (Inbarani et al., 2014a). Feature reduction (also called as variable selection, attribute selection or subset selection) is the process of choosing a prevailing subset of features that is most correlated to the decision classes (Inbarani et al., 2014b; Wahhab, 2015). The digital format of the medical images gives an opportunity for further analysis that may lead to a more accurate diagnosis and helps in the optimized patient management.
In media image processing, it is very difficult to cope as data volumes are increasingly growing. Big data analysis currently plays a crucial role in the management, organization and analysis of data. Big data solutions provide new outlines for the analysis of medical images in a manner close to big data. It is time to develop new methods/architectures based on big data technologies for the complete processing of biomedical image data. In biomedical image processing, images are recorded in various imaging modes such as CT Scan, MRI, X-Ray, Ultra Sound, and PET-CT. It geenrates a large volume of images and it is very difficult for the radiologist to manage these large volumes of images. This research is an attempt to apply digital image processing and machine learning techniques in the area of medical image analysis and recognition. It focuses on developing a method to segment and diagnose the Acute Lymphoblastic Leukemia (ALL) nucleus. Fuzzy Clustering is utilized to segment the nucleus. It is a novel population-based stochastic algorithm proposed by Civicoglu (Civicioglu, 2013). After segmenting the image, relevant and representative features are extracted from the segmented nucleus. During this process, different kinds of features are extracted, namely, shape, colour and texture features.