Herawati, Yoshi Inne
Unknown Affiliation

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

A Data Pre-processing Strategy Utilizing Adaptive Masking for the Classification of Pediatric Pneumonia Using VGG-16 Herawati, Yoshi Inne; Rahmat, Basuki; Hendra Maulana
Jurnal Sistem Informasi dan Ilmu Komputer Vol. 8 No. 1 (2024): JUSIKOM: JURNAL SISTEM INFROMASI ILMU KOMPUTER
Publisher : Fakultas Teknologi dan Ilmu Komputer Universitas Prima Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34012/jurnalsisteminformasidanilmukomputer.v8i1.5604

Abstract

Pneumonia is still a leading cause of death in children, especially in areas with limited medical resources. This study aims to test several pre-processes to find the best set of pre-processes that can be applied to the children's chest X-ray dataset by applying adaptive masking, histogram equalization, CLAHE and Gaussian blur. Then, childhood pneumonia is classified using a CNN architecture, namely VGG-16. By applying these pre-processing methods, this study is divided into several scenarios. The highest accuracy was obtained from scenario 1, which used a combination of adaptive masking, histogram equalization and Gaussian blur, resulting in an accuracy of 94%. Scenario 2 uses histogram equalization and Gaussian blur with an accuracy of 92%. Then Scenario 3 uses a histogram equalization replacement for CLAHE with a combination of adaptive masking, CLAHE and Gaussian blur with 93% accuracy. Finally, scenario 4 uses a combination of CLAHE and Gaussian blur methods with 91% accuracy. In addition, this research also addresses the challenges posed by unbalanced data sets and the need for highly accurate detection tools.