Anggrainy Togi Marito Siregar, Dina Jumiatul Fitri, Happy Alyzhya Haay, Rasi Kasim Samosir

CLASSIFICATION OF PNEUMONIA DISEASE USING THE MINI XCEPTION MODEL ON X-RAY DATA

Introduction

Classification of pneumonia disease using the mini xception model on x-ray data. Classify bacterial, viral, and normal pneumonia from X-ray data using the Mini Xception model. Achieved 86% overall accuracy, with strong precision & sensitivity for bacterial cases.

Abstract

Pneumonia is a disease that often causes death in Indonesia. In general, many that cause a person to develop pneumonia include pneumonia due to bacterial, viral, mycoplasma pneumonia, fungal pneumonia. There are many ways to detect a patient grouped into one type of pneumonia. One way is to use an X-ray machine. X-ray is technology that can send waves of electromagnetic radiation briefly to scan the condition of the inside of the body. In this study, we tried to classify patients affected by bacterial pneumonia and viral pneumonia as well as normal people. The data we use is a picture of the lungs taken from the X-ray results. This research was conducted by applying the mini Xception model using the python program. The model can predict the results of X-ray scans that belong to the class of bacterial pneumonia very well, as seen from the value of precision and sensitivity of 80 and 97 percent, respectively. Viral pneumonia class can not be predicted as good as the two previous classes, but the results obtained are quite good as seen from the value of precision and sensitivity of 85 and 67 percent, respectively. The overall accuracy of the model obtained is 0.86.

Full Text

Comments

You need to be logged in to post a comment.