Pulmonary rontgen classification to detect pneumonia disease using convolutional neural networks

Zuherman Rustam, Rivan Pratama Yuda, Hamimah Alatas, Chelvian Aroef

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Every organism is known to have different structural and biological system, specifically in human immunity. If the immune system weakens, the body is susceptible to disease especially pneumonia disease. Pneumonia disease is caused by the bacterium Streptococcus pneumonia, and according to the World Health Organization (WHO), it is identified as the leading cause of death in children worldwide, which is about 16%, for those under the age of 5. Meanwhile, someone who is predicted to have pneumonia by a doctor is recommended for an X-ray. Convolutional neural networks (CNNs) is an accurate method to help the doctor's predicted correctly. CNNs is divided into two important parts, feature extraction layer (convolutional layer and pooling layer) and fully connected layer. CNNs method is commonly used for image data classification. Therefore, CNNs is suitable to classify pneumonia based on lung X-ray in order to obtain accurate prediction results. And then, the results can be seen based on the graph of the accuracy value and the loss value. When CNNs method applied on the dataset, an accuracy rate of 97% was obtained. Based on accuracy rate, it shows that CNNs can be applied to image data (especially lung X-ray) for classification of pneumonia disease.

Original languageEnglish
Pages (from-to)1522-1528
Number of pages7
JournalTelkomnika (Telecommunication Computing Electronics and Control)
Volume18
Issue number3
DOIs
Publication statusPublished - 1 Jan 2020

Keywords

  • Classification
  • CNNs
  • Pneumonia
  • X-ray

Fingerprint Dive into the research topics of 'Pulmonary rontgen classification to detect pneumonia disease using convolutional neural networks'. Together they form a unique fingerprint.

  • Cite this