Backpropagation Performance Against Support Vector Machine in Detecting Tuberculosis Based on Lung X-Ray Image

DOI

10.2991/icmeme-18.2019.19How to use a DOI?

Keywords

artificial neural network, backpropagation, accuracy, detection, tuberculosis

Abstract

Tuberculosis (TB) is known as an infectious disease caused by bacterium Mycobacterium Tuberculosis. It is one of the highest diseases that occur in Indonesia. The lung disease can be identified by analyzing the x-ray image of the lung. The problem that followed is that the x-ray images were analyzed separately by the specialist physician at separate times, so the patient should consult a doctor after getting the x-ray image. In this study, we create a modeling design that can detect TB disease early by using artificial neural network method that is backpropagation by using Matlab Software, furthermore analyze the performance of the modeling based on the level of accuracy. In training process this system uses 441 images while for the test used 221 x-ray images. The system’s phases were started with preprocessing including median filter process and histogram equalization to improve image quality. The results of preprocessing is then classified with Backpropagation algorithm through training process. The results showed that TBC detection system can be built using backpropagation method with 4400 hidden layer hidden neurons with accuracy of 81.45% from the test process result. The accuracy of NN Backpropagation is better than SVM method whose accuracy reaches of 78.73%.

Copyright

© 2019, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Arizal
AU  - Andani Achmad
AU  - Andini Dani Achmad
PY  - 2019/03
DA  - 2019/03
TI  - Backpropagation Performance Against Support Vector Machine in Detecting Tuberculosis Based on Lung X-Ray Image
BT  - Proceedings of the First International Conference on Materials Engineering and Management - Engineering Section (ICMEMe 2018)
PB  - Atlantis Press
SP  - 84
EP  - 88
SN  - 2352-5401
UR  - https://doi.org/10.2991/icmeme-18.2019.19
DO  - 10.2991/icmeme-18.2019.19
ID  - 2019/03
ER  -