Aulia, Suci and Hadiyoso, Sugondo (2022) Tuberculosis Detection in X-Ray Image Using Deep Learning Approach with VGG-16 Architecture. [Artikel Dosen]
![[thumbnail of Tuberculosis Detection in X-Ray Image Using Deep Learning Approach with VGG-16 Architecture.pdf]](http://eprints.uad.ac.id/style/images/fileicons/text.png) |
Text
Tuberculosis Detection in X-Ray Image Using Deep Learning Approach with VGG-16 Architecture.pdf Download (515kB) |
Abstract
Tuberculosis (TB) is a chronic disease still the main problem in Indonesia. However, this disease can be cured with drugs at a particular time after the patient is detected as having TB. TB diagnosis or screening can be made through x-ray imaging of the chest cavity by a radiology specialist. The Mantoux test can then be used to confirm the diagnosis. X-ray images often have varying contrasts that lead to true negatives or false negatives. Whereas generally, a chest x-ray is the initial examination of TB. Error detection will have a fatal impact on treatment therapy. Therefore, this study proposed a system for TB detection based on x-ray images using deep learning. The system developed uses a Convolutional Neural Network (CNN) with the VGG-16 architecture. In the performance test stage, 700 normal and 140 TB chest x-ray images were used. The simulation results show that the proposed system can classify normal and TB lungs with an accuracy of 99.76%. The highest accuracy is achieved using batch size=50. This system is expected to assist radiology in detecting tuberculosis on X-Ray images of the lungs. The contribution of this study is to build a machine learning model for TB detection and optimization of model parameters to get the best accuracy.
| Item Type: | Artikel Dosen |
|---|---|
| Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering |
| Divisi / Prodi: | Faculty of Industrial Technology (Fakultas Teknologi Industri) > S1-Electrical Engineering (S1-Teknik Elektro) |
| Depositing User: | M.Eng. Alfian Ma'arif |
| Date Deposited: | 24 Jan 2023 06:34 |
| Last Modified: | 24 Jan 2023 06:34 |
| URI: | http://eprints.uad.ac.id/id/eprint/37451 |
Actions (login required)
 |
View Item |