Deteksi Pneumonia Dengan Menggunakan Deep Learning Faster Region Convolutional Neural Network (Faster R-CNN)

Hafidz Daffa, Hekmatyar (2022) Deteksi Pneumonia Dengan Menggunakan Deep Learning Faster Region Convolutional Neural Network (Faster R-CNN). Undergraduate Thesis thesis, Institut Teknologi Telkom Purwokerto.

	Text Cover.pdf Download (1MB)
	Text Abstract.pdf Download (223kB)
	Text Abstrak.pdf Download (672kB)
	Text BAB I.pdf Download (265kB)
	Text BAB II.pdf Download (1MB)
	Text BAB III.pdf Download (901kB)
	Text BAB IV.pdf Restricted to Registered users only Download (934kB) | Request a copy
	Text BAB V.pdf Download (226kB)
	Text Daftar Pustaka.pdf Download (351kB)
	Text Lampiran.pdf Restricted to Registered users only Download (3MB) | Request a copy

Abstract

Pneumonia is a lung infection involving alveoli (air sacs) caused by microbes, including bacteria, viruses, or fungi that can cause inflammation in bronchioles and alveoli that close the capillaries. Pneumonia is a very dangerous disease if not treated quickly. COVID-19 virus (coronavirus disease 2019) is a new virus that was discovered in the city of Wuhan, Hubei Province, China at the end of 2019, the COVID-19 virus caused various health problems, one of which was pneumonia. With increasing number of positive COVID-19 patients, pneumonia detection becomes very important in detecting presence pneumonia early and quickly in COVID-19 patients, to minimize the symptoms caused by pneumonia due to COVID-19 virus. Detection of radiological images in early stages is very difficult, tedious and time consuming for radiologist, screening multiple scans carefully takes a lot of time. From these problems, authors create a model to determine the location of pneumonia by using x-ray images to accelerate handling COVID-19 patients using Faster Region Convolutional Neural Network (Faster R-CNN) algorithm. This research that was made only up to manufacture of pneumonia detection models. The research conducted, the author uses CNN VGG16 and ResNet50 architecture as framework for modeling. From two architectural tests carried out, VGG16 model gave the highest mAP (mean Average Precision) level of 14.1% while ResNet50 is 11.9%. Then for tests carried out with x-ray images of pneumonia caused by COVID-19 virus, VGG16 model was able to predict the location of pneumonia. COVID19 with highest success rate of 86% while ResNet50 is 84%. Keywords: Pneumonia, COVID-19, Faster R-CNN

Actions (login required)

View Item