A Hybridized Pre-Processing Method for Detecting Tuberculosis using Deep Learning

• Main Authors: Elashmawy, A.M.A., Elamvazuthi, I., Ali, S.S.A., Natarajan, E., Paramasivam, S.
• Format: Conference or Workshop Item
• Published: Institute of Electrical and Electronics Engineers Inc. 2021
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85124166674&doi=10.1109%2fICIAS49414.2021.9642622&partnerID=40&md5=50cae21bf4b0058f2a9ead326e851537; http://eprints.utp.edu.my/29184/

Tuberculosis (TB), a disease that targets the individual's lungs and can cause fatalities can be cured if detected and treated early. Computer Aided Diagnosis (CAD) systems could be utilized to detect the presence of TB in Chest X-Ray Images (CXR). This paper proposes to investigate a hybridized pre-processing method for Convolutional Neural Network (CNN) CAD system for detecting TB in CXR images. The aim of this research is to improve the performance of CNNs by combining two different pre-processing methods and to further multi-classify different manifestation of TB. In this research, the experimental design is to apply augmentation and segmentation to CXR images as pre-processing and use a pretrained CNN model to classify the pre-processed images. It is hypothesized that the research would improve the accuracy and Area Under Curve (AUC) of detection of TB in CXR images. © 2021 IEEE.