Skripsi

KLASIFIKASI PENYAKIT COVID-19 MELALUI HASIL CITRA X-RAY MENGGUNAKAN METODE CNN

---

Penilaian

---

Covid-19 is a disease caused by a corona virus. The test to detect COVID-19 is PCR (Real-time polymerase chain reaction) and radiological images such as CT-Scan and X-Ray also have an important role in detecting COVID-19 patients at an early stage. Many researchers in the IT field observe that X-Ray images can be utilized and developed to help detect COVID-19 in patients. One of the methods used in radiological image classification is the Convolutional Neural Network (CNN). This study aims to obtain performance results from CNN for detecting COVID-19 and Normal disease in X-Ray images with datasets originating from hospitals in Indonesia using 4 architectures, namely ResNet50, MobileNet, VGG19, and Modification Model. The training results show that the model with the MobileNet architecture provides the best performance with an accuracy of 95.31% after going through 500 epochs. The time for the process is 7 hours 30 minutes. This model is then used to test 400 data with a success rate of 81%. This model was then tested on medical personnel with new data and the results obtained were 70% or able to detect 7 out of 10 new X-ray image data. The use of the Convolutional Neural Network (CNN) method has proven effective in detecting COVID-19 through chest X-Ray images.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2307002560	T116070	T1160702023	Central Library (Referens)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T1160702023

Publisher

Indralaya : Prodi Teknik Elektro, Fakultas Teknik Universitas Sriwijaya., 2023

Collation

xiv, 45 hlm.; Ilus.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

539.722 207

Content Type

-

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Teknik Elektro
Sinar Gamma
Sinar X

Specific Detail Info

-

Statement of Responsibility

PITRIA

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment