Learning from the Scratch for Tuberculosis (TB) Bacilli Detection Using DSOD

Abstract

Tuberculosis (TB) is an infectious disease caused by the micro-bacteria Tuberculosis which can be transmitted through sputum sprinkling. Tuberculosis is a direct infectious disease caused by tuberculosis bacilli. Most tuberculosis bacilli attack the lungs but It could also be other organs of the body. Several studies that have been conducted previously aimed to reduce the burden of observing Tuberculosis bacilli using the processing method digital image. As technology continues to develop, the methods used to process digital images are also increasing. In this study, researchers used Deep Learning in digital image processing. This method aims to make the system or machine able to recognize tuberculosis bacilli with various kinds of data that have been shown. Deep Learning Method trains the machine to recognize Tuberculosis bacilli through learning various images of bacilli under varying conditions so that the machine can recognize the tray under various conditions slides. We believe that a fast and reliable diagnostic aid system with accurate information to help medical workers in the field is necessary.