Big Data: Scientific Study and Analysis of Medical Image Data Using X-Rays with Feature Extraction and Classification Techniques

Abstract

As medical imaging technology advances, the range of images to be collected, processed, recorded, preserved and recovered at medical centers is exponentially increasing. In the recent years classification and the retrieval of medical images have thus been a subject of popularity. Several systematic methods are used for evaluating and classifying vast X-ray images that have a crucial impact on clinical practice. Here we categorize four types of x-rays images: head, hand, leg, and chest. This paper presents Local Binary Pattern (LBP), Local Directional Pattern (LDP), and Histogram of Oriented Gradients (HOG) for the feature extraction and Support Vector Machine (SVM) for feature classification. The dataset are created from open source and Kaggle dataset .Experiments are conducted on these datasets and best result is obtained using HOG and Linear SVM, which produces 98.09 of accuracy.